linux-kernel - Re: [PATCH R4 6/7] mm: Extend memory hotplug API to allow memory hotplug in virtual guests

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]

Date:	Thu, 10 Mar 2011 09:51:36 +0100
From:	Daniel Kiper <dkiper@...-space.pl>
To:	Dave Hansen <dave@...ux.vnet.ibm.com>
Cc:	Daniel Kiper <dkiper@...-space.pl>, ian.campbell@...rix.com,
	akpm@...ux-foundation.org, andi.kleen@...el.com,
	haicheng.li@...ux.intel.com, fengguang.wu@...el.com,
	jeremy@...p.org, konrad.wilk@...cle.com,
	dan.magenheimer@...cle.com, v.tolstov@...fip.ru, pasik@....fi,
	wdauchy@...il.com, rientjes@...gle.com,
	xen-devel@...ts.xensource.com, linux-kernel@...r.kernel.org,
	linux-mm@...ck.org
Subject: Re: [PATCH R4 6/7] mm: Extend memory hotplug API to allow memory hotplug in virtual guests

On Tue, Mar 08, 2011 at 03:51:12PM -0800, Dave Hansen wrote:
> On Tue, 2011-03-08 at 22:50 +0100, Daniel Kiper wrote:
> > +int add_virtual_memory(u64 *size)
> > +{
> > +	int nid;
> > +	u64 start;
> > +
> > +	start = PFN_PHYS(SECTION_ALIGN(max_pfn));
> > +	*size = (((*size >> PAGE_SHIFT) & PAGE_SECTION_MASK) + PAGES_PER_SECTION) << PAGE_SHIFT;
>
> Why use PFN_PHYS() in one case but not the other?

I know that this is the same, however, I think PFN_PHYS() usage suggest
that I do a PFN/address manipulation. It is not true in that case (I do
an operation on region size) and I would like to avoid that ambiguity.

> I'd also highly suggest using the ALIGN() macro in cases like this.  It
> makes it much more readable:

OK.

> 	*size = PFN_PHYS(ALIGN(*size, SECTION_SIZE)));
>
> > +	nid = memory_add_physaddr_to_nid(start);
> > +
> > +	return add_memory(nid, start, *size);
> > +}
>
> Could you talk a little bit more about how 'size' gets used?  Also, are
> we sure we want an interface where we're so liberal with 'size'?  It
> seems like requiring that it be section-aligned is a fair burden to
> place on the caller.  That way, we're not in a position of _guessing_
> what the caller wants (aligning up or down).

I do not have like this function since I created it. However,
I decided to sent it for review. It does not simplify anything
(add_memory() as a generic function is sufficient) and it is
too inflexible. Now, I am sure that everything in its body
should be moved to platform specific module (in that case Xen).
I am going to that on next patch release.

Daniel
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/