linux-kernel - Re: [PATCH 3/3] cma: move allocation from HIGHMEM to a helper function

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <aGa9mworOcOE6GhV@kernel.org>
Date: Thu, 3 Jul 2025 20:27:55 +0300
From: Mike Rapoport <rppt@...nel.org>
To: Oscar Salvador <osalvador@...e.de>
Cc: Andrew Morton <akpm@...ux-foundation.org>,
	Alexandre Ghiti <alexghiti@...osinc.com>,
	David Hildenbrand <david@...hat.com>,
	Pratyush Yadav <ptyadav@...zon.de>, linux-kernel@...r.kernel.org,
	linux-mm@...ck.org
Subject: Re: [PATCH 3/3] cma: move allocation from HIGHMEM to a helper
 function

On Thu, Jul 03, 2025 at 11:53:06AM +0200, Oscar Salvador wrote:
> On Wed, Jul 02, 2025 at 08:36:05PM +0300, Mike Rapoport wrote:
> > From: "Mike Rapoport (Microsoft)" <rppt@...nel.org>
> > 
> > When CONFIG_HIGMEM is enabled, __cma_declare_contiguous_nid() first
> > tries to allocate the area from HIGHMEM and if that fails it falls back
> > to allocation from low memory.
> > 
> > Split allocation from HIGMEM into a helper function to further decouple
> > logic related to CONFIG_HIGHMEM.
> > 
> > Signed-off-by: Mike Rapoport (Microsoft) <rppt@...nel.org>
> > ---
> >  mm/cma.c | 52 +++++++++++++++++++++++++++++-----------------------
> >  1 file changed, 29 insertions(+), 23 deletions(-)
> > 
> > diff --git a/mm/cma.c b/mm/cma.c
> > index 1df8ff312d99..0a24c46f3296 100644
> > --- a/mm/cma.c
> > +++ b/mm/cma.c
> > @@ -376,6 +376,30 @@ static int __init cma_fixed_reserve(phys_addr_t base, phys_addr_t size)
> >  	return 0;
> >  }
> >  
> > +static phys_addr_t __init cma_alloc_highmem(phys_addr_t base, phys_addr_t size,
> > +			phys_addr_t align, phys_addr_t *limit, int nid)
> > +{
> > +	phys_addr_t addr = 0;
> > +
> > +	if (IS_ENABLED(CONFIG_HIGHMEM)) {
> > +		phys_addr_t highmem = __pa(high_memory - 1) + 1;
> > +
> > +		/*
> > +		 * All pages in the reserved area must come from the same zone.
> > +		 * If the requested region crosses the low/high memory boundary,
> > +		 * try allocating from high memory first and fall back to low
> > +		 * memory in case of failure.
> > +		 */
> > +		if (base < highmem && *limit > highmem) {
> > +			addr = memblock_alloc_range_nid(size, align, highmem,
> > +							*limit, nid, true);
> > +			*limit = highmem;
> > +		}
> > +	}
> 
> Not a big deal, but maybe better to do it in one function? Maybe even move
> the CONFIG_PHYS_ADDR_T_64BIT block in there as well? So memblock_alloc_range_nid()
> calls would be contained in one place and the X86_64/HIGHMEM comments as
> well.
> Just a thought.

Yeah, this will be neater, thanks!
Will send v2 shortly.

> -- 
> Oscar Salvador
> SUSE Labs

-- 
Sincerely yours,
Mike.