Re: mm/filemap.c:3307 do_sync_mmap_readahead() warn: assigning (-12) to unsigned variable 'ra->order'

[Dan Carpenter]

On Tue, Jan 13, 2026 at 01:24:20PM +0100, Jan Kara wrote:
> On Tue 13-01-26 14:12:28, Dan Carpenter wrote:
> > tree:   https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git master
> > head:   b71e635feefc852405b14620a7fc58c4c80c0f73
> > commit: 9f1edf1aedac1b287355f63f768ba4275de72dca mm: readahead: make thp readahead conditional to mmap_miss logic
> > config: x86_64-randconfig-161-20260113 (https://download.01.org/0day-ci/archive/20260113/202601131311.g3lyUAks-lkp@xxxxxxxxx/config)
                                            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

> > compiler: clang version 20.1.8 (https://github.com/llvm/llvm-project 87f0227cb60147a26a1eeb4fb06e3b505e9c7261)
> > smatch version: v0.5.0-8985-g2614ff1a
> > 
> > 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@xxxxxxxxx>
> > | Reported-by: Dan Carpenter <dan.carpenter@xxxxxxxxxx>
> > | Closes: https://lore.kernel.org/r/202601131311.g3lyUAks-lkp@xxxxxxxxx/
> > 
> > smatch warnings:
> > mm/filemap.c:3307 do_sync_mmap_readahead() warn: assigning (-12) to unsigned variable 'ra->order'
> ...
> > 9f1edf1aedac1b2 Roman Gushchin          2025-10-06  3306  		ra->async_size = HPAGE_PMD_NR;
> > 9f1edf1aedac1b2 Roman Gushchin          2025-10-06 @3307  		ra->order = HPAGE_PMD_ORDER;
> > 
> > HPAGE_PMD_SHIFT is zero in this config?
> > ra->order is a u16.
> 
> What's the config exactly? I suppose CONFIG_TRANSPARENT_HUGEPAGE is
> disabled in this case, isn't it? In that case force_thp_readahead will
> never be true and thus the branch with the assignment can never be reached
> AFAICT...
> 

Ah, yes.  You are right.  force_thp_readahead is false in this
config.

The truth is, I could easily disable these kinds of warnings in smatch
because this is marked as an impossible path.  if (is_impossible_path())
return...  I've done that on a case by case basis in the past but maybe I
should just do it globally.

regards,
dan carpenter