Re: [PATCH 4.4 15/50] s390/alternative: use a copy of the facility bit mask

From: Martin Schwidefsky <schwidefsky@de.ibm.com>
To: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Jiri Slaby <jslaby@suse.cz>,
	linux-kernel@vger.kernel.org, stable@vger.kernel.org,
	David Hildenbrand <david@redhat.com>,
	Cornelia Huck <cohuck@redhat.com>
Subject: Re: [PATCH 4.4 15/50] s390/alternative: use a copy of the facility bit mask
Date: Mon, 7 May 2018 08:07:07 +0200	[thread overview]
Message-ID: <20180507080707.2a41ae28@mschwideX1> (raw)
In-Reply-To: <20180504221808.GB24813@kroah.com>

On Fri, 4 May 2018 15:18:08 -0700
Greg Kroah-Hartman <gregkh@linuxfoundation.org> wrote:

> On Fri, May 04, 2018 at 09:37:20AM +0200, Jiri Slaby wrote:
> > On 04/27/2018, 03:58 PM, Greg Kroah-Hartman wrote:  
> > > 4.4-stable review patch.  If anyone has any objections, please let me know.
> > > 
> > > ------------------
> > > 
> > > From: Martin Schwidefsky <schwidefsky@de.ibm.com>
> > > 
> > > 
> > > [ Upstream commit cf1489984641369611556bf00c48f945c77bcf02 ]
> > > 
> > > To be able to switch off specific CPU alternatives with kernel parameters
> > > make a copy of the facility bit mask provided by STFLE and use the copy
> > > for the decision to apply an alternative.  
> > ...  
> > > --- a/arch/s390/include/asm/facility.h
> > > +++ b/arch/s390/include/asm/facility.h
> > > @@ -13,6 +13,24 @@
> > >  
> > >  #define MAX_FACILITY_BIT (256*8)	/* stfle_fac_list has 256 bytes */  
> > 
> > I wonder if the below (plus __test_facility) is correct in 4.4, given
> > MAX_FACILITY_BIT is defined as such and not as sizeof(stfle_fac_list *
> > 8) as in upstream?  
> 
> I'm going to defer to Marin here, as he did the backport...
> Martin?

Good catch. With MAX_FACILITY_BIT == 2048 and the patch applied the result
for a test_facility/__test_facility call with a facility number >= 1024
would give an incorrect result. Fortunately there are no such calls in
the current 4.4 kernel source. And there are no facilities defined with
bit numbers this large, so even out-of-tree code would not do this if
it is sane.

To correct this the MAX_FACILITY_BIT define needs to be reduced to 1024
which would require the patch pointed out be Heiko:

commit 6f5165e864d240d15675cc2fb5a369d57e1f60d0
Author: Heiko Carstens <heiko.carstens@de.ibm.com>
Date:   Mon Mar 20 14:29:50 2017 +0100

    s390/facilites: use stfle_fac_list array size for MAX_FACILITY_BIT

I would say yes, it *does* make sense to include this patch even if it
does not fix anything.

-- 
blue skies,
   Martin.

"Reality continues to ruin my life." - Calvin.