[PATCH linux-next] riscv/mm: fix up the mm-fixes merge of free_pmd_table()

[PATCH linux-next] riscv/mm: fix up the mm-fixes merge of free_pmd_table()

[Sasha Levin]

The mm-fixes merge reuses an older resolution of the conflict in
remove_pud_mapping() and keeps the free_pmd_table() call that passes
is_vmemmap. Commit f998a21c2ba5 ("riscv/mm: fix hotplug page-table
destructor handling") in mm-fixes removed that argument, so a riscv
build with CONFIG_MEMORY_HOTPLUG fails:

  arch/riscv/mm/init.c:1657:35: error: too many arguments to function call, expected 2, have 3

Drop the argument to match the mm-fixes version of the call.

Fixes: f383866da287 ("Merge branch 'for-next-fixes' of https://git.kernel.org/pub/scm/linux/kernel/git/mm/linux.git")
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
Mark, this applies on top of next-20261030 and is meant to be folded into
the mm-fixes merge resolution. Next/merge.log says it was resolved using
a previous resolution, I think the one from the risc-v/mm conflict on
2026-10-12, which had the same problem (next-20261012 has it too).

Build tested with clang: riscv defconfig plus MEMORY_HOTPLUG and
MEMORY_HOTREMOVE, W=1 build of arch/riscv/mm/ (also mm/memory_hotplug.o and
mm/sparse-vmemmap.o), and plain defconfig.

 arch/riscv/mm/init.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/riscv/mm/init.c b/arch/riscv/mm/init.c
index 2400537a01215..e822fce0af4b0 100644
--- a/arch/riscv/mm/init.c
+++ b/arch/riscv/mm/init.c
@@ -1654,7 +1654,7 @@ static void __meminit remove_pud_mapping(pud_t *pud_base, unsigned long addr, un
 		remove_pmd_mapping(pmd_base, addr, next, is_vmemmap, altmap);
 
 		if (pgtable_l4_enabled())
-			free_pmd_table(pmd_base, pudp, is_vmemmap);
+			free_pmd_table(pmd_base, pudp);
 	}
 }
 

base-commit: 7649ba73f9e7e1e04d0b46213e39c79ae09dd55c
-- 
2.53.0

Re: [PATCH linux-next] riscv/mm: fix up the mm-fixes merge of free_pmd_table()

[Andrew Morton]

On Fri, 30 Oct 2026 18:14:01 -0400 Sasha Levin <sashal@kernel.org> wrote:

> The mm-fixes merge reuses an older resolution of the conflict in
> remove_pud_mapping() and keeps the free_pmd_table() call that passes
> is_vmemmap. Commit f998a21c2ba5 ("riscv/mm: fix hotplug page-table
> destructor handling") in mm-fixes removed that argument, so a riscv
> build with CONFIG_MEMORY_HOTPLUG fails:
> 
>   arch/riscv/mm/init.c:1657:35: error: too many arguments to function call, expected 2, have 3
> 
> Drop the argument to match the mm-fixes version of the call.
> 
> ---
> Mark, this applies on top of next-20261030 and is meant to be folded into
> the mm-fixes merge resolution. Next/merge.log says it was resolved using
> a previous resolution, I think the one from the risc-v/mm conflict on
> 2026-10-12, which had the same problem (next-20261012 has it too).

Thanks.  The "mm: Fix page-table teardown during memory hot-remove", v3
series was problematic, but I now have all four patches lined up in
Muchun's as-sent order in mm-hotfixes-unstable.  I'll push that in the
next hour or three.  Then hopefully any -next merge resolution can just
go away.