All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Roger Pau Monné" <roger.pau@citrix.com>
To: Elias El Yandouzi <eliasely@amazon.com>
Cc: xen-devel@lists.xenproject.org, julien@xen.org,
	pdurrant@amazon.com, dwmw@amazon.com,
	Hongyan Xia <hongyxia@amazon.com>,
	Andrew Cooper <andrew.cooper3@citrix.com>,
	George Dunlap <george.dunlap@citrix.com>,
	Jan Beulich <jbeulich@suse.com>,
	Stefano Stabellini <sstabellini@kernel.org>,
	Julien Grall <jgrall@amazon.com>
Subject: Re: [PATCH V3 (resend) 12/19] x86/setup: vmap heap nodes when they are outside the direct map
Date: Tue, 14 May 2024 17:02:45 +0200	[thread overview]
Message-ID: <ZkN9FWP3VMZS-s-y@macbook> (raw)
In-Reply-To: <20240513134046.82605-13-eliasely@amazon.com>

On Mon, May 13, 2024 at 01:40:39PM +0000, Elias El Yandouzi wrote:
> From: Hongyan Xia <hongyxia@amazon.com>
> 
> When we do not have a direct map, archs_mfn_in_direct_map() will always
> return false, thus init_node_heap() will allocate xenheap pages from an
> existing node for the metadata of a new node. This means that the
> metadata of a new node is in a different node, slowing down heap
> allocation.
> 
> Since we now have early vmap, vmap the metadata locally in the new node.
> 
> Signed-off-by: Hongyan Xia <hongyxia@amazon.com>
> Signed-off-by: Julien Grall <jgrall@amazon.com>
> Signed-off-by: Elias El Yandouzi <eliasely@amazon.com>
> 
> ----
> 
>     Changes in v2:
>         * vmap_contig_pages() was renamed to vmap_contig()
>         * Fix indentation and coding style
> 
>     Changes from Hongyan's version:
>         * arch_mfn_in_direct_map() was renamed to
>           arch_mfns_in_direct_map()
>         * Use vmap_contig_pages() rather than __vmap(...).
>         * Add missing include (xen/vmap.h) so it compiles on Arm
> 
> diff --git a/xen/common/page_alloc.c b/xen/common/page_alloc.c
> index dfb2c05322..3c0909f333 100644
> --- a/xen/common/page_alloc.c
> +++ b/xen/common/page_alloc.c
> @@ -136,6 +136,7 @@
>  #include <xen/sched.h>
>  #include <xen/softirq.h>
>  #include <xen/spinlock.h>
> +#include <xen/vmap.h>
>  
>  #include <asm/flushtlb.h>
>  #include <asm/page.h>
> @@ -605,22 +606,44 @@ static unsigned long init_node_heap(int node, unsigned long mfn,
>          needed = 0;
>      }
>      else if ( *use_tail && nr >= needed &&
> -              arch_mfns_in_directmap(mfn + nr - needed, needed) &&
>                (!xenheap_bits ||
> -               !((mfn + nr - 1) >> (xenheap_bits - PAGE_SHIFT))) )
> +              !((mfn + nr - 1) >> (xenheap_bits - PAGE_SHIFT))) )

Unrelated change?  The indentation was correct for this line and you
are breaking it.

>      {
> -        _heap[node] = mfn_to_virt(mfn + nr - needed);
> -        avail[node] = mfn_to_virt(mfn + nr - 1) +
> -                      PAGE_SIZE - sizeof(**avail) * NR_ZONES;
> +        if ( arch_mfns_in_directmap(mfn + nr - needed, needed) )
> +        {
> +            _heap[node] = mfn_to_virt(mfn + nr - needed);
> +            avail[node] = mfn_to_virt(mfn + nr - 1) +
> +                          PAGE_SIZE - sizeof(**avail) * NR_ZONES;
> +        }
> +        else
> +        {
> +            mfn_t needed_start = _mfn(mfn + nr - needed);
> +
> +            _heap[node] = vmap_contig(needed_start, needed);
> +            BUG_ON(!_heap[node]);
> +            avail[node] = (void *)(_heap[node]) + (needed << PAGE_SHIFT) -
> +                          sizeof(**avail) * NR_ZONES;
> +        }

You could shorten the blocks I think:

if ( arch_mfns_in_directmap(mfn + nr - needed, needed) )
    _heap[node] = mfn_to_virt(mfn + nr - needed);
else
    _heap[node] = vmap_contig(_mfn(mfn + nr - needed), needed);

BUG_ON(!_heap[node]);
avail[node] = (void *)(_heap[node]) + (needed << PAGE_SHIFT) -
                  sizeof(**avail) * NR_ZONES;

So that more part of the logic is shared between both.

>      }
>      else if ( nr >= needed &&
> -              arch_mfns_in_directmap(mfn, needed) &&
>                (!xenheap_bits ||
> -               !((mfn + needed - 1) >> (xenheap_bits - PAGE_SHIFT))) )
> +              !((mfn + needed - 1) >> (xenheap_bits - PAGE_SHIFT))) )
>      {
> -        _heap[node] = mfn_to_virt(mfn);
> -        avail[node] = mfn_to_virt(mfn + needed - 1) +
> -                      PAGE_SIZE - sizeof(**avail) * NR_ZONES;
> +        if ( arch_mfns_in_directmap(mfn, needed) )
> +        {
> +            _heap[node] = mfn_to_virt(mfn);
> +            avail[node] = mfn_to_virt(mfn + needed - 1) +
> +                          PAGE_SIZE - sizeof(**avail) * NR_ZONES;
> +        }
> +        else
> +        {
> +            mfn_t needed_start = _mfn(mfn);
> +
> +            _heap[node] = vmap_contig(needed_start, needed);
> +            BUG_ON(!_heap[node]);
> +            avail[node] = (void *)(_heap[node]) + (needed << PAGE_SHIFT) -
> +                          sizeof(**avail) * NR_ZONES;
> +        }

Same here.

Thanks, Roger.


  reply	other threads:[~2024-05-14 15:03 UTC|newest]

Thread overview: 70+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-05-13 13:40 [PATCH V3 (resend) 00/19] Remove the directmap Elias El Yandouzi
2024-05-13 13:40 ` [PATCH V3 (resend) 01/19] x86: Create per-domain mapping of guest_root_pt Elias El Yandouzi
2024-05-14 14:51   ` Jan Beulich
2024-05-15 18:25     ` Elias El Yandouzi
2024-05-16  7:17       ` Jan Beulich
2024-06-13 16:31         ` Elias El Yandouzi
2024-06-14  6:23           ` Jan Beulich
2024-06-17  7:33             ` Roger Pau Monné
2024-05-13 13:40 ` [PATCH V3 (resend) 02/19] x86/pv: Domheap pages should be mapped while relocating initrd Elias El Yandouzi
2024-05-13 15:40   ` Roger Pau Monné
2024-05-13 13:40 ` [PATCH V3 (resend) 03/19] x86/pv: Rewrite how building PV dom0 handles domheap mappings Elias El Yandouzi
2024-05-13 16:49   ` Roger Pau Monné
2024-05-14 14:58     ` Jan Beulich
2024-05-14 15:03   ` Jan Beulich
2024-07-16 16:12     ` Elias El Yandouzi
2024-07-17 10:45       ` Jan Beulich
2024-05-13 13:40 ` [PATCH V3 (resend) 04/19] x86: Lift mapcache variable to the arch level Elias El Yandouzi
2024-05-14  8:21   ` Roger Pau Monné
2024-05-15 13:11   ` Jan Beulich
2024-07-16 17:06   ` Alejandro Vallejo
2024-07-17 12:41     ` Alejandro Vallejo
2024-05-13 13:40 ` [PATCH V3 (resend) 05/19] x86/mapcache: Initialise the mapcache for the idle domain Elias El Yandouzi
2024-05-14  8:42   ` Roger Pau Monné
2024-05-15 13:44     ` Jan Beulich
2024-05-13 13:40 ` [PATCH V3 (resend) 06/19] x86: Add a boot option to enable and disable the direct map Elias El Yandouzi
2024-05-14  9:20   ` Roger Pau Monné
2024-05-14 10:20     ` Roger Pau Monné
2024-05-15 13:54     ` Jan Beulich
2024-05-16  9:19       ` Roger Pau Monné
2024-05-16  9:24         ` Jan Beulich
2024-05-15 13:59   ` Jan Beulich
2024-05-15 16:02   ` Jan Beulich
2024-05-13 13:40 ` [PATCH V3 (resend) 07/19] xen/x86: Add support for the PMAP Elias El Yandouzi
2024-05-14  9:40   ` Roger Pau Monné
2024-05-14  9:43     ` Jan Beulich
2024-05-14 10:22       ` Roger Pau Monné
2024-05-14 10:26         ` Jan Beulich
2024-05-14 11:51           ` Roger Pau Monné
2024-05-14 12:33             ` Jan Beulich
2024-05-13 13:40 ` [PATCH V3 (resend) 08/19] xen/x86: Add build assertion for fixmap entries Elias El Yandouzi
2024-05-14  9:42   ` Roger Pau Monné
2024-05-14  9:45     ` Jan Beulich
2024-05-15 14:03   ` Jan Beulich
2024-05-13 13:40 ` [PATCH V3 (resend) 09/19] x86/domain_page: Remove the fast paths when mfn is not in the directmap Elias El Yandouzi
2024-05-14 11:48   ` Roger Pau Monné
2024-05-15 14:21   ` Jan Beulich
2024-05-13 13:40 ` [PATCH V3 (resend) 10/19] xen/page_alloc: Add a path for xenheap when there is no direct map Elias El Yandouzi
2024-05-14 13:07   ` Roger Pau Monné
2024-05-15 15:13   ` Jan Beulich
2024-05-13 13:40 ` [PATCH V3 (resend) 11/19] x86/setup: Leave early boot slightly earlier Elias El Yandouzi
2024-05-14 14:11   ` Roger Pau Monné
2024-05-15 15:22   ` Jan Beulich
2024-05-13 13:40 ` [PATCH V3 (resend) 12/19] x86/setup: vmap heap nodes when they are outside the direct map Elias El Yandouzi
2024-05-14 15:02   ` Roger Pau Monné [this message]
2024-05-15 15:28   ` Jan Beulich
2024-05-13 13:40 ` [PATCH V3 (resend) 13/19] x86/setup: Do not create valid mappings when directmap=no Elias El Yandouzi
2024-05-14 15:39   ` Roger Pau Monné
2024-05-15 15:50     ` Jan Beulich
2024-05-15 15:59   ` Jan Beulich
2024-05-13 13:40 ` [PATCH V3 (resend) 14/19] Rename mfn_to_virt() calls Elias El Yandouzi
2024-05-14 15:45   ` Roger Pau Monné
2024-05-14 16:22     ` Jan Beulich
2024-05-15  9:38       ` Roger Pau Monné
2024-05-15  9:42         ` Jan Beulich
2024-05-16  8:57   ` Jan Beulich
2024-05-13 13:40 ` [PATCH V3 (resend) 15/19] Rename maddr_to_virt() calls Elias El Yandouzi
2024-05-13 13:40 ` [PATCH V3 (resend) 16/19] xen/arm32: mm: Rename 'first' to 'root' in init_secondary_pagetables() Elias El Yandouzi
2024-05-13 13:40 ` [PATCH V3 (resend) 17/19] xen/arm64: mm: Use per-pCPU page-tables Elias El Yandouzi
2024-05-13 13:40 ` [PATCH V3 (resend) 18/19] xen/arm64: Implement a mapcache for arm64 Elias El Yandouzi
2024-05-13 13:40 ` [PATCH V3 (resend) 19/19] xen/arm64: Allow the admin to enable/disable the directmap Elias El Yandouzi

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=ZkN9FWP3VMZS-s-y@macbook \
    --to=roger.pau@citrix.com \
    --cc=andrew.cooper3@citrix.com \
    --cc=dwmw@amazon.com \
    --cc=eliasely@amazon.com \
    --cc=george.dunlap@citrix.com \
    --cc=hongyxia@amazon.com \
    --cc=jbeulich@suse.com \
    --cc=jgrall@amazon.com \
    --cc=julien@xen.org \
    --cc=pdurrant@amazon.com \
    --cc=sstabellini@kernel.org \
    --cc=xen-devel@lists.xenproject.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.