From: "Kirill A. Shutemov" <kirill@shutemov.name>
To: Dave Hansen <dave.hansen@intel.com>
Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>,
Ingo Molnar <mingo@redhat.com>,
x86@kernel.org, Thomas Gleixner <tglx@linutronix.de>,
"H. Peter Anvin" <hpa@zytor.com>,
Tom Lendacky <thomas.lendacky@amd.com>,
Kai Huang <kai.huang@linux.intel.com>,
Jacob Pan <jacob.jun.pan@linux.intel.com>,
linux-kernel@vger.kernel.org, linux-mm@kvack.org
Subject: Re: [PATCHv5 12/19] x86/mm: Implement prep_encrypted_page() and arch_free_page()
Date: Mon, 23 Jul 2018 12:50:40 +0300 [thread overview]
Message-ID: <20180723095040.w67jp7c7cnxezuwp@kshutemo-mobl1> (raw)
In-Reply-To: <a05d800e-4c18-88e0-388c-093fc3dac6ec@intel.com>
On Wed, Jul 18, 2018 at 04:53:27PM -0700, Dave Hansen wrote:
> The description doesn't mention the potential performance implications
> of this patch. That's criminal at this point.
>
> > --- a/arch/x86/mm/mktme.c
> > +++ b/arch/x86/mm/mktme.c
> > @@ -1,4 +1,5 @@
> > #include <linux/mm.h>
> > +#include <linux/highmem.h>
> > #include <asm/mktme.h>
> >
> > phys_addr_t mktme_keyid_mask;
> > @@ -49,3 +50,51 @@ int vma_keyid(struct vm_area_struct *vma)
> > prot = pgprot_val(vma->vm_page_prot);
> > return (prot & mktme_keyid_mask) >> mktme_keyid_shift;
> > }
> > +
> > +void prep_encrypted_page(struct page *page, int order, int keyid, bool zero)
> > +{
> > + int i;
> > +
> > + /* It's not encrypted page: nothing to do */
> > + if (!keyid)
> > + return;
>
> prep_encrypted_page() is called in the fast path in the page allocator.
> This out-of-line copy costs a function call for all users and this is
> also out of the reach of the compiler to understand that keyid!=0 is
> unlikely.
>
> I think this needs to be treated to the inline-in-the-header treatment.
Okay. Again as a macros.
> > + /*
> > + * The hardware/CPU does not enforce coherency between mappings of the
> > + * same physical page with different KeyIDs or encryption keys.
> > + * We are responsible for cache management.
> > + *
> > + * We flush cache before allocating encrypted page
> > + */
> > + clflush_cache_range(page_address(page), PAGE_SIZE << order);
>
> It's also worth pointing out that this must be done on the keyid alias
> direct map, not the normal one.
>
> Wait a sec... How do we know which direct map to use?
page_address() -> lowmem_page_address() -> page_to_virt()
page_to_virt() returns virtual address from the right direct mapping.
> > + for (i = 0; i < (1 << order); i++) {
> > + /* All pages coming out of the allocator should have KeyID 0 */
> > + WARN_ON_ONCE(lookup_page_ext(page)->keyid);
> > + lookup_page_ext(page)->keyid = keyid;
> > +
> > + /* Clear the page after the KeyID is set. */
> > + if (zero)
> > + clear_highpage(page);
> > +
> > + page++;
> > + }
> > +}
> > +
> > +void arch_free_page(struct page *page, int order)
> > +{
> > + int i;
> > +
> > + /* It's not encrypted page: nothing to do */
> > + if (!page_keyid(page))
> > + return;
>
> Ditto on pushing this to a header.
>
> > + clflush_cache_range(page_address(page), PAGE_SIZE << order);
>
> OK, how do we know which copy of the direct map to use, here?
The same way as above.
> > + for (i = 0; i < (1 << order); i++) {
> > + /* Check if the page has reasonable KeyID */
> > + WARN_ON_ONCE(lookup_page_ext(page)->keyid > mktme_nr_keyids);
> > + lookup_page_ext(page)->keyid = 0;
> > + page++;
> > + }
> > +}
> >
>
--
Kirill A. Shutemov
next prev parent reply other threads:[~2018-07-23 9:50 UTC|newest]
Thread overview: 73+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-07-17 11:20 [PATCHv5 00/19] MKTME enabling Kirill A. Shutemov
2018-07-17 11:20 ` [PATCHv5 01/19] mm: Do no merge VMAs with different encryption KeyIDs Kirill A. Shutemov
2018-07-17 11:20 ` [PATCHv5 02/19] mm: Do not use zero page in encrypted pages Kirill A. Shutemov
2018-07-18 17:36 ` Dave Hansen
2018-07-19 7:16 ` Kirill A. Shutemov
2018-07-19 13:58 ` Dave Hansen
2018-07-20 12:16 ` Kirill A. Shutemov
2018-07-17 11:20 ` [PATCHv5 03/19] mm/ksm: Do not merge pages with different KeyIDs Kirill A. Shutemov
2018-07-18 17:38 ` Dave Hansen
2018-07-19 7:32 ` Kirill A. Shutemov
2018-07-19 14:02 ` Dave Hansen
2018-07-20 12:23 ` Kirill A. Shutemov
2018-07-17 11:20 ` [PATCHv5 04/19] mm/page_alloc: Unify alloc_hugepage_vma() Kirill A. Shutemov
2018-07-18 17:43 ` Dave Hansen
2018-07-17 11:20 ` [PATCHv5 05/19] mm/page_alloc: Handle allocation for encrypted memory Kirill A. Shutemov
2018-07-18 23:03 ` Dave Hansen
2018-07-19 8:27 ` Kirill A. Shutemov
2018-07-19 14:05 ` Dave Hansen
2018-07-20 12:25 ` Kirill A. Shutemov
2018-07-26 14:25 ` Michal Hocko
2018-07-17 11:20 ` [PATCHv5 06/19] mm/khugepaged: Handle encrypted pages Kirill A. Shutemov
2018-07-18 23:11 ` Dave Hansen
2018-07-19 8:59 ` Kirill A. Shutemov
2018-07-19 14:13 ` Dave Hansen
2018-07-20 12:29 ` Kirill A. Shutemov
2018-07-17 11:20 ` [PATCHv5 07/19] x86/mm: Mask out KeyID bits from page table entry pfn Kirill A. Shutemov
2018-07-18 23:13 ` Dave Hansen
2018-07-19 9:54 ` Kirill A. Shutemov
2018-07-19 14:19 ` Dave Hansen
2018-07-20 12:31 ` Kirill A. Shutemov
2018-07-17 11:20 ` [PATCHv5 08/19] x86/mm: Introduce variables to store number, shift and mask of KeyIDs Kirill A. Shutemov
2018-07-18 23:19 ` Dave Hansen
2018-07-19 10:21 ` Kirill A. Shutemov
2018-07-19 12:37 ` Thomas Gleixner
2018-07-19 13:12 ` Kirill A. Shutemov
2018-07-19 13:18 ` Thomas Gleixner
2018-07-19 13:23 ` Kirill A. Shutemov
2018-07-19 13:40 ` Thomas Gleixner
2018-07-20 12:34 ` Kirill A. Shutemov
2018-07-20 13:17 ` Thomas Gleixner
2018-07-20 13:40 ` Kirill A. Shutemov
2018-07-19 14:23 ` Dave Hansen
2018-07-20 12:34 ` Kirill A. Shutemov
2018-07-31 0:08 ` Kai Huang
2018-07-17 11:20 ` [PATCHv5 09/19] x86/mm: Preserve KeyID on pte_modify() and pgprot_modify() Kirill A. Shutemov
2018-07-18 23:30 ` Dave Hansen
2018-07-20 12:42 ` Kirill A. Shutemov
2018-07-17 11:20 ` [PATCHv5 10/19] x86/mm: Implement page_keyid() using page_ext Kirill A. Shutemov
2018-07-18 23:38 ` Dave Hansen
2018-07-23 9:45 ` Kirill A. Shutemov
2018-07-23 17:22 ` Alison Schofield
2018-07-17 11:20 ` [PATCHv5 11/19] x86/mm: Implement vma_keyid() Kirill A. Shutemov
2018-07-18 23:40 ` Dave Hansen
2018-07-23 9:47 ` Kirill A. Shutemov
2018-07-17 11:20 ` [PATCHv5 12/19] x86/mm: Implement prep_encrypted_page() and arch_free_page() Kirill A. Shutemov
2018-07-18 23:53 ` Dave Hansen
2018-07-23 9:50 ` Kirill A. Shutemov [this message]
2018-07-17 11:20 ` [PATCHv5 13/19] x86/mm: Rename CONFIG_RANDOMIZE_MEMORY_PHYSICAL_PADDING Kirill A. Shutemov
2018-07-17 11:20 ` [PATCHv5 14/19] x86/mm: Allow to disable MKTME after enumeration Kirill A. Shutemov
2018-07-17 11:20 ` [PATCHv5 15/19] x86/mm: Detect MKTME early Kirill A. Shutemov
2018-07-17 11:20 ` [PATCHv5 16/19] x86/mm: Calculate direct mapping size Kirill A. Shutemov
2018-07-17 11:20 ` [PATCHv5 17/19] x86/mm: Implement sync_direct_mapping() Kirill A. Shutemov
2018-07-19 0:01 ` Dave Hansen
2018-07-23 10:04 ` Kirill A. Shutemov
2018-07-23 12:25 ` Dave Hansen
2018-07-17 11:20 ` [PATCHv5 18/19] x86/mm: Handle encrypted memory in page_to_virt() and __pa() Kirill A. Shutemov
2018-07-18 22:21 ` Thomas Gleixner
2018-07-23 10:12 ` Kirill A. Shutemov
2018-07-26 17:26 ` Dave Hansen
2018-07-27 13:49 ` Kirill A. Shutemov
2018-07-17 11:20 ` [PATCHv5 19/19] x86: Introduce CONFIG_X86_INTEL_MKTME Kirill A. Shutemov
2018-08-15 7:48 ` Pavel Machek
2018-08-17 9:24 ` Kirill A. Shutemov
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20180723095040.w67jp7c7cnxezuwp@kshutemo-mobl1 \
--to=kirill@shutemov.name \
--cc=dave.hansen@intel.com \
--cc=hpa@zytor.com \
--cc=jacob.jun.pan@linux.intel.com \
--cc=kai.huang@linux.intel.com \
--cc=kirill.shutemov@linux.intel.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mingo@redhat.com \
--cc=tglx@linutronix.de \
--cc=thomas.lendacky@amd.com \
--cc=x86@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).