Re: [linux-next:master 12265/12944] arch/arm64/include/asm/pgtable.h:904:16: error: implicit declaration of function 'pud_leaf'; did you mean 'pmd_leaf'?

[Peter Xu <peterx@redhat.com>]

On Sun, Mar 10, 2024 at 04:10:37PM +0800, kernel test robot wrote:
> tree:   https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git master
> head:   8ffc8b1bbd505e27e2c8439d326b6059c906c9dd
> commit: 924bd6a8c96767a05323d575bdefd664631dce73 [12265/12944] mm/x86: drop two unnecessary pud_leaf() definitions
> config: arm64-randconfig-r031-20230409 (https://download.01.org/0day-ci/archive/20240310/202403101607.a42gaLOS-lkp@intel.com/config)
> compiler: aarch64-linux-gcc (GCC) 13.2.0
> reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20240310/202403101607.a42gaLOS-lkp@intel.com/reproduce)
> 
> If you fix the issue in a separate patch/commit (i.e. not just a new version of
> the same patch/commit), kindly add following tags
> | Reported-by: kernel test robot <lkp@intel.com>
> | Closes: https://lore.kernel.org/oe-kbuild-all/202403101607.a42gaLOS-lkp@intel.com/
> 
> All errors (new ones prefixed by >>):
> 
>    In file included from include/linux/pgtable.h:6,
>                     from arch/arm64/include/asm/io.h:12,
>                     from include/linux/io.h:13,
>                     from include/acpi/acpi_io.h:5,
>                     from include/linux/acpi.h:39,
>                     from include/acpi/apei.h:9,
>                     from include/acpi/ghes.h:5,
>                     from include/linux/arm_sdei.h:8,
>                     from arch/arm64/kernel/asm-offsets.c:10:
>    arch/arm64/include/asm/pgtable.h: In function 'pud_user_accessible_page':
> >> arch/arm64/include/asm/pgtable.h:904:16: error: implicit declaration of function 'pud_leaf'; did you mean 'pmd_leaf'? [-Werror=implicit-function-declaration]
>      904 |         return pud_leaf(pud) && (pud_user(pud) || pud_user_exec(pud));
>          |                ^~~~~~~~
>          |                pmd_leaf
>    cc1: some warnings being treated as errors
>    make[3]: *** [scripts/Makefile.build:116: arch/arm64/kernel/asm-offsets.s] Error 1
>    make[3]: Target 'prepare' not remade because of errors.
>    make[2]: *** [Makefile:1199: prepare0] Error 2
>    make[2]: Target 'prepare' not remade because of errors.
>    make[1]: *** [Makefile:240: __sub-make] Error 2
>    make[1]: Target 'prepare' not remade because of errors.
>    make: *** [Makefile:240: __sub-make] Error 2
>    make: Target 'prepare' not remade because of errors.
> 
> 
> vim +904 arch/arm64/include/asm/pgtable.h
> 
> ed928a3402d8a2 Tong Tiangen 2022-05-17  901  
> ed928a3402d8a2 Tong Tiangen 2022-05-17  902  static inline bool pud_user_accessible_page(pud_t pud)
> ed928a3402d8a2 Tong Tiangen 2022-05-17  903  {
> 730a11f982e61a Liu Shixin   2022-11-22 @904  	return pud_leaf(pud) && (pud_user(pud) || pud_user_exec(pud));
> ed928a3402d8a2 Tong Tiangen 2022-05-17  905  }
> ed928a3402d8a2 Tong Tiangen 2022-05-17  906  #endif
> ed928a3402d8a2 Tong Tiangen 2022-05-17  907  
> 
> :::::: The code at line 904 was first introduced by commit
> :::::: 730a11f982e61aaef758ab552dfb7c30de79e99b arm64/mm: add pud_user_exec() check in pud_user_accessible_page()
> 
> :::::: TO: Liu Shixin <liushixin2@huawei.com>
> :::::: CC: Will Deacon <will@kernel.org>
> 
> -- 
> 0-DAY CI Kernel Test Service
> https://github.com/intel/lkp-tests/wiki
> 

There's another possibly relevant report for riscv:

https://lore.kernel.org/r/202403090900.OwPqmRuI-lkp@intel.com

I think I messed up the nopmd use case.  Since this commit already landed
mm-stable, I assume there's no way I provide a fixup.

I attached a formal patch below, but I don't know how to test it myself,
and I'm not 100% confident.  Before I post a formal patch, can anyone let
me know how I can kickoff the test bot to test the patch (or help me do the
kickoff)?

Thanks,

======8<=======
From acc5598c7af0b7c9a486a49b8c562bb351e8d35b Mon Sep 17 00:00:00 2001
From: Peter Xu <peterx@redhat.com>
Date: Mon, 11 Mar 2024 12:10:45 -0400
Subject: [PATCH] mm: Recover pud_leaf() definitions in nopmd case

This reverts one change in commit 924bd6a8c96767a05323d575bdefd664631dce73.

One issue with that is it broke nopmd builds for at least both arm64 and
riscv (CONFIG_PGTABLE_LEVELS=2).  The other issue is it was overlooked that
it's a common change rather than x86 specific (relevant to the commit
message of the commit).

Normally there's no need for empty definition of pXd_leaf() because of the
fallback functions, however this logic may not apply to pgtable-nopmd.h,
because that's a header that can even be used by arch *pgtable.h headers,
which can use the *_leaf() definitions _before_ the fallback functions are
defined.  Leave it there to pass PGTABLE_LEVELS=2 builds.

Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202403090900.OwPqmRuI-lkp@intel.com/
Closes: https://lore.kernel.org/oe-kbuild-all/202403101607.a42gaLOS-lkp@intel.com/
Fixes: 924bd6a8c967 ("mm/x86: drop two unnecessary pud_leaf() definitions")
Signed-off-by: Peter Xu <peterx@redhat.com>
---
 include/asm-generic/pgtable-nopmd.h | 1 +
 1 file changed, 1 insertion(+)

diff --git a/include/asm-generic/pgtable-nopmd.h b/include/asm-generic/pgtable-nopmd.h
index fa27e16bbe1b..8ffd64e7a24c 100644
--- a/include/asm-generic/pgtable-nopmd.h
+++ b/include/asm-generic/pgtable-nopmd.h
@@ -31,6 +31,7 @@ static inline int pud_none(pud_t pud)		{ return 0; }
 static inline int pud_bad(pud_t pud)		{ return 0; }
 static inline int pud_present(pud_t pud)	{ return 1; }
 static inline int pud_user(pud_t pud)		{ return 0; }
+static inline int pud_leaf(pud_t pud)		{ return 0; }
 static inline void pud_clear(pud_t *pud)	{ }
 #define pmd_ERROR(pmd)				(pud_ERROR((pmd).pud))
 
-- 
2.44.0