Re: [akpm-mm:mm-stable 72/140] arch/sparc/mm/init_64.c:2582:16: error: implicit declaration of function 'vmemmap_pgd_populate'; did you mean 'vmemmap_populate'?

[Philip Li <philip.li@intel.com>]

On Tue, Mar 24, 2026 at 09:26:58PM -0700, Andrew Morton wrote:
> On Wed, 25 Mar 2026 11:57:27 +0800 Philip Li <philip.li@intel.com> wrote:
> 
> > > The coincidence is too great.  I assume Kernel Test Robot is somehow
> > > treating the mm-stable differently from other branches?  Can we please
> > > not do that?
> > 
> > Sorry for this "coincidence", the bot does treat all branches equally and
> > there's no special handling of mm-stable. One possibility is the randconfig
> > is just generated on 20260325 "sparc-randconfig-001-20260325", which is able
> > to expose this issue for sparc arch.
> 
> huh, OK, thanks.  I was able to reproduce it with sparc64 defconfig.

Apologize for my conclusion, i did a check of internal data when replies mail at
that time and only find the errors of randconfig. Since sparc64-defconfig is recorded
as done but not shown in the error record, I made the wrong conclusion that this is
exposed by randconfig. The truth is sparc64-defconfig does cause build failure.

recent_errors/
└── sparc-randconfig-001-20260325
    ├── arch-sparc-mm-init_64.c:error:implicit-declaration-of-function-vmemmap_p4d_populate
    ├── arch-sparc-mm-init_64.c:error:implicit-declaration-of-function-vmemmap_pgd_populate
    ├── arch-sparc-mm-init_64.c:error:implicit-declaration-of-function-vmemmap_pud_populate
    ├── arch-sparc-mm-init_64.c:warning:assignment-to-p4d_t-aka-struct-anonymous-from-int-makes-pointer-from-integer-without-a-cast
    ├── arch-sparc-mm-init_64.c:warning:assignment-to-pud_t-aka-struct-anonymous-from-int-makes-pointer-from-integer-without-a-cast
    └── arch-sparc-mm-init_64.c:warning:initialization-of-pgd_t-aka-struct-anonymous-from-int-makes-pointer-from-integer-without-a-cast

I will do more investigation to understand why bot doesn't report issue out during the
7 weeks window, and fix the bug. Sorry again about this miss.

Thanks

> 
>