linux-kernel - Re: [PATCH] s390/sclp: Add check for get_zeroed

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: <20250217161637.21424Ce0-hca@linux.ibm.com>
Date: Mon, 17 Feb 2025 17:16:37 +0100
From: Heiko Carstens <hca@...ux.ibm.com>
To: Heiko Carstens <hca@...ux.ibm.com>
Cc: Haoxiang Li <haoxiang_li2024@....com>, gor@...ux.ibm.com,
        agordeev@...ux.ibm.com, borntraeger@...ux.ibm.com, svens@...ux.ibm.com,
        schwidefsky@...ibm.com, linux-s390@...r.kernel.org,
        linux-kernel@...r.kernel.org, stable@...r.kernel.org
Subject: Re: [PATCH] s390/sclp: Add check for get_zeroed_page()

On Mon, Feb 17, 2025 at 05:01:17PM +0100, Heiko Carstens wrote:
> On Mon, Feb 17, 2025 at 11:31:46PM +0800, Haoxiang Li wrote:
> > diff --git a/drivers/s390/char/sclp_con.c b/drivers/s390/char/sclp_con.c
> > index e5d947c763ea..7447076b1ec1 100644
> > --- a/drivers/s390/char/sclp_con.c
> > +++ b/drivers/s390/char/sclp_con.c
> > @@ -282,6 +282,8 @@ sclp_console_init(void)
> >  	/* Allocate pages for output buffering */
> >  	for (i = 0; i < sclp_console_pages; i++) {
> >  		page = (void *) get_zeroed_page(GFP_KERNEL | GFP_DMA);
> > +		if (!page)
> > +			return -ENOMEM;
> >  		list_add_tail(page, &sclp_con_pages);
> 
> We can add this check, however if this early allocation would fail a
> null pointer dereference would be the last problem we would have to
> think about.
> 
> Anyway:
> Acked-by: Heiko Carstens <hca@...ux.ibm.com>

Wait, I take that back. Now I think I remember why I didn't add error
handling back then: the above exit would also indicate a potential
memory leak, since this is a loop allocating several pages; so all
already allocated pages must be freed, which would ask for even more
completely pointless error handling.

This is very early code where any allocation failure would lead to a
crash in any case. So either do the full exercise or we leave the code
as it is.