linux-kernel - Re: [PATCH] mm/early_ioremap: Print the starting physical address in __early

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: <20260112032907.GA60059@k08j02272.eu95sqa>
Date: Mon, 12 Jan 2026 11:29:07 +0800
From: Hou Wenlong <houwenlong.hwl@...group.com>
To: Mike Rapoport <rppt@...nel.org>
Cc: Andrew Morton <akpm@...ux-foundation.org>, linux-kernel@...r.kernel.org,
	David Hildenbrand <david@...nel.org>,
	Lorenzo Stoakes <lorenzo.stoakes@...cle.com>,
	"Liam R. Howlett" <Liam.Howlett@...cle.com>,
	Vlastimil Babka <vbabka@...e.cz>,
	Suren Baghdasaryan <surenb@...gle.com>,
	Michal Hocko <mhocko@...e.com>, linux-mm@...ck.org
Subject: Re: [PATCH] mm/early_ioremap: Print the starting physical address in
 __early_ioremap()

On Sun, Jan 11, 2026 at 11:15:45AM +0200, Mike Rapoport wrote:
> On Sat, Jan 10, 2026 at 03:57:21PM -0800, Andrew Morton wrote:
> > On Fri,  9 Jan 2026 21:31:51 +0800 Hou Wenlong <houwenlong.hwl@...group.com> wrote:
> > 
> > > The debug WARN() printing occurs after the while loop, so the
> > > 'phys_addr' reflects the last physical address rather than the actual
> > > starting physical address, which is not useful for debugging. To
> > > simplify, the WARN() statement could be moved up before the loop instead
> > > of introducing a new variable to record the original 'phys_addr' value.
> > > Additionally, swap the print order of 'slot_virt[slot]' and 'offset', as
> > > this will enhance output readability.
> > > 
> > 
> > yep, thanks.
> > 
> > > --- a/mm/early_ioremap.c
> > > +++ b/mm/early_ioremap.c
> > > @@ -139,6 +139,9 @@ __early_ioremap(resource_size_t phys_addr, unsigned long size, pgprot_t prot)
> > >  	if (WARN_ON(nrpages > NR_FIX_BTMAPS))
> > >  		return NULL;
> > >  
> > > +	WARN(early_ioremap_debug, "%s(%pa, %08lx) [%d] => %08lx + %08lx\n",
> > > +	     __func__, &phys_addr, size, slot, slot_virt[slot], offset);
> > > +
> > >  	/*
> > >  	 * Ok, go for it..
> > >  	 */
> > > @@ -152,8 +155,6 @@ __early_ioremap(resource_size_t phys_addr, unsigned long size, pgprot_t prot)
> > >  		--idx;
> > >  		--nrpages;
> > >  	}
> > > -	WARN(early_ioremap_debug, "%s(%pa, %08lx) [%d] => %08lx + %08lx\n",
> > > -	     __func__, &phys_addr, size, slot, offset, slot_virt[slot]);
> > 
> > It's strange that this code uses WARN when nothing is wrong.  The rest
> > of the function uses WARN appropriately.  But that's off-topic!
> 
> It seems an historic thing. A nice cleanup would be to replace
> WARN(early_ioremap_debug) with pr_warn() + dump_stack().
>
Yes, I see that the initial code uses WARN(), so I didn't change it, as
cleanup is a separate concern and may not be well-received. However, I
can make the change as you suggested.

Thanks!

> > Reviewed-by: Andrew Morton <akpm@...ux-foundation.org>
> > 
> 
> -- 
> Sincerely yours,
> Mike.