linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Peter Zijlstra <peterz@infradead.org>
To: Suren Baghdasaryan <surenb@google.com>
Cc: akpm@linux-foundation.org, willy@infradead.org,
	liam.howlett@oracle.com, lorenzo.stoakes@oracle.com,
	david.laight.linux@gmail.com, mhocko@suse.com, vbabka@suse.cz,
	hannes@cmpxchg.org, mjguzik@gmail.com, oliver.sang@intel.com,
	mgorman@techsingularity.net, david@redhat.com, peterx@redhat.com,
	oleg@redhat.com, dave@stgolabs.net, paulmck@kernel.org,
	brauner@kernel.org, dhowells@redhat.com, hdanton@sina.com,
	hughd@google.com, lokeshgidra@google.com, minchan@google.com,
	jannh@google.com, shakeel.butt@linux.dev, souravpanda@google.com,
	pasha.tatashin@soleen.com, klarasmodin@gmail.com,
	richard.weiyang@gmail.com, corbet@lwn.net,
	linux-doc@vger.kernel.org, linux-mm@kvack.org,
	linux-kernel@vger.kernel.org, kernel-team@android.com
Subject: Re: [PATCH v9 12/17] mm: move lesser used vma_area_struct members into the last cacheline
Date: Wed, 15 Jan 2025 11:50:56 +0100	[thread overview]
Message-ID: <20250115105056.GY5388@noisy.programming.kicks-ass.net> (raw)
In-Reply-To: <20250111042604.3230628-13-surenb@google.com>

On Fri, Jan 10, 2025 at 08:25:59PM -0800, Suren Baghdasaryan wrote:
> Move several vma_area_struct members which are rarely or never used
> during page fault handling into the last cacheline to better pack
> vm_area_struct. As a result vm_area_struct will fit into 3 as opposed
> to 4 cachelines. New typical vm_area_struct layout:
> 
> struct vm_area_struct {
>     union {
>         struct {
>             long unsigned int vm_start;              /*     0     8 */
>             long unsigned int vm_end;                /*     8     8 */
>         };                                           /*     0    16 */
>         freeptr_t          vm_freeptr;               /*     0     8 */
>     };                                               /*     0    16 */
>     struct mm_struct *         vm_mm;                /*    16     8 */
>     pgprot_t                   vm_page_prot;         /*    24     8 */
>     union {
>         const vm_flags_t   vm_flags;                 /*    32     8 */
>         vm_flags_t         __vm_flags;               /*    32     8 */
>     };                                               /*    32     8 */
>     unsigned int               vm_lock_seq;          /*    40     4 */

Does it not make sense to move this seq field near the refcnt?

>     /* XXX 4 bytes hole, try to pack */
> 
>     struct list_head           anon_vma_chain;       /*    48    16 */
>     /* --- cacheline 1 boundary (64 bytes) --- */
>     struct anon_vma *          anon_vma;             /*    64     8 */
>     const struct vm_operations_struct  * vm_ops;     /*    72     8 */
>     long unsigned int          vm_pgoff;             /*    80     8 */
>     struct file *              vm_file;              /*    88     8 */
>     void *                     vm_private_data;      /*    96     8 */
>     atomic_long_t              swap_readahead_info;  /*   104     8 */
>     struct mempolicy *         vm_policy;            /*   112     8 */
>     struct vma_numab_state *   numab_state;          /*   120     8 */
>     /* --- cacheline 2 boundary (128 bytes) --- */
>     refcount_t          vm_refcnt (__aligned__(64)); /*   128     4 */
> 
>     /* XXX 4 bytes hole, try to pack */
> 
>     struct {
>         struct rb_node     rb (__aligned__(8));      /*   136    24 */
>         long unsigned int  rb_subtree_last;          /*   160     8 */
>     } __attribute__((__aligned__(8))) shared;        /*   136    32 */
>     struct anon_vma_name *     anon_name;            /*   168     8 */
>     struct vm_userfaultfd_ctx  vm_userfaultfd_ctx;   /*   176     8 */
> 
>     /* size: 192, cachelines: 3, members: 18 */
>     /* sum members: 176, holes: 2, sum holes: 8 */
>     /* padding: 8 */
>     /* forced alignments: 2, forced holes: 1, sum forced holes: 4 */
> } __attribute__((__aligned__(64)));




  parent reply	other threads:[~2025-01-15 10:51 UTC|newest]

Thread overview: 140+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-01-11  4:25 [PATCH v9 00/17] reimplement per-vma lock as a refcount Suren Baghdasaryan
2025-01-11  4:25 ` [PATCH v9 01/17] mm: introduce vma_start_read_locked{_nested} helpers Suren Baghdasaryan
2025-01-11  4:25 ` [PATCH v9 02/17] mm: move per-vma lock into vm_area_struct Suren Baghdasaryan
2025-01-11  4:25 ` [PATCH v9 03/17] mm: mark vma as detached until it's added into vma tree Suren Baghdasaryan
2025-01-11  4:25 ` [PATCH v9 04/17] mm: introduce vma_iter_store_attached() to use with attached vmas Suren Baghdasaryan
2025-01-13 11:58   ` Lorenzo Stoakes
2025-01-13 16:31     ` Suren Baghdasaryan
2025-01-13 16:44       ` Lorenzo Stoakes
2025-01-13 16:47       ` Lorenzo Stoakes
2025-01-13 19:09         ` Suren Baghdasaryan
2025-01-14 11:38           ` Lorenzo Stoakes
2025-01-11  4:25 ` [PATCH v9 05/17] mm: mark vmas detached upon exit Suren Baghdasaryan
2025-01-13 12:05   ` Lorenzo Stoakes
2025-01-13 17:02     ` Suren Baghdasaryan
2025-01-13 17:13       ` Lorenzo Stoakes
2025-01-13 19:11         ` Suren Baghdasaryan
2025-01-13 20:32           ` Vlastimil Babka
2025-01-13 20:42             ` Suren Baghdasaryan
2025-01-14 11:36               ` Lorenzo Stoakes
2025-01-11  4:25 ` [PATCH v9 06/17] types: move struct rcuwait into types.h Suren Baghdasaryan
2025-01-13 14:46   ` Lorenzo Stoakes
2025-01-11  4:25 ` [PATCH v9 07/17] mm: allow vma_start_read_locked/vma_start_read_locked_nested to fail Suren Baghdasaryan
2025-01-13 15:25   ` Lorenzo Stoakes
2025-01-13 17:53     ` Suren Baghdasaryan
2025-01-14 11:48       ` Lorenzo Stoakes
2025-01-11  4:25 ` [PATCH v9 08/17] mm: move mmap_init_lock() out of the header file Suren Baghdasaryan
2025-01-13 15:27   ` Lorenzo Stoakes
2025-01-13 17:53     ` Suren Baghdasaryan
2025-01-11  4:25 ` [PATCH v9 09/17] mm: uninline the main body of vma_start_write() Suren Baghdasaryan
2025-01-13 15:52   ` Lorenzo Stoakes
2025-01-11  4:25 ` [PATCH v9 10/17] refcount: introduce __refcount_{add|inc}_not_zero_limited Suren Baghdasaryan
2025-01-11  6:31   ` Hillf Danton
2025-01-11  9:59     ` Suren Baghdasaryan
2025-01-11 10:00       ` Suren Baghdasaryan
2025-01-11 12:13       ` Hillf Danton
2025-01-11 17:11         ` Suren Baghdasaryan
2025-01-11 23:44           ` Hillf Danton
2025-01-12  0:31             ` Suren Baghdasaryan
2025-01-15  9:39           ` Peter Zijlstra
2025-01-16 10:52             ` Hillf Danton
2025-01-11 12:39   ` David Laight
2025-01-11 17:07     ` Matthew Wilcox
2025-01-11 18:30     ` Paul E. McKenney
2025-01-11 22:19       ` David Laight
2025-01-11 22:50         ` [PATCH v9 10/17] refcount: introduce __refcount_{add|inc}_not_zero_limited - clang 17.0.1 bug David Laight
2025-01-12 11:37           ` David Laight
2025-01-12 17:56             ` Paul E. McKenney
2025-01-11  4:25 ` [PATCH v9 11/17] mm: replace vm_lock and detached flag with a reference count Suren Baghdasaryan
2025-01-11 11:24   ` Mateusz Guzik
2025-01-11 20:14     ` Suren Baghdasaryan
2025-01-11 20:16       ` Suren Baghdasaryan
2025-01-11 20:31       ` Mateusz Guzik
2025-01-11 20:58         ` Suren Baghdasaryan
2025-01-11 20:38       ` Vlastimil Babka
2025-01-13  1:47       ` Wei Yang
2025-01-13  2:25         ` Wei Yang
2025-01-13 21:14           ` Suren Baghdasaryan
2025-01-13 21:08         ` Suren Baghdasaryan
2025-01-15 10:48       ` Peter Zijlstra
2025-01-15 11:13         ` Peter Zijlstra
2025-01-15 15:00           ` Suren Baghdasaryan
2025-01-15 15:35             ` Peter Zijlstra
2025-01-15 15:38               ` Peter Zijlstra
2025-01-15 16:22                 ` Suren Baghdasaryan
2025-01-15 16:00           ` [PATCH] refcount: Strengthen inc_not_zero() Peter Zijlstra
2025-01-16 15:12             ` Suren Baghdasaryan
2025-01-17 15:41             ` Will Deacon
2025-01-27 14:09               ` Will Deacon
2025-01-27 19:21                 ` Suren Baghdasaryan
2025-01-28 23:51                   ` Suren Baghdasaryan
2025-02-06  2:52                     ` [PATCH 1/1] refcount: provide ops for cases when object's memory can be reused Suren Baghdasaryan
2025-02-06 10:41                       ` Vlastimil Babka
2025-02-06  3:03                     ` [PATCH] refcount: Strengthen inc_not_zero() Suren Baghdasaryan
2025-02-13 23:04                       ` Suren Baghdasaryan
2025-01-17 16:13             ` Matthew Wilcox
2025-01-12  2:59   ` [PATCH v9 11/17] mm: replace vm_lock and detached flag with a reference count Wei Yang
2025-01-12 17:35     ` Suren Baghdasaryan
2025-01-13  0:59       ` Wei Yang
2025-01-13  2:37   ` Wei Yang
2025-01-13 21:16     ` Suren Baghdasaryan
2025-01-13  9:36   ` Wei Yang
2025-01-13 21:18     ` Suren Baghdasaryan
2025-01-15  2:58   ` Wei Yang
2025-01-15  3:12     ` Suren Baghdasaryan
2025-01-15 12:05       ` Wei Yang
2025-01-15 15:01         ` Suren Baghdasaryan
2025-01-16  1:37           ` Wei Yang
2025-01-16  1:41             ` Suren Baghdasaryan
2025-01-16  9:10               ` Wei Yang
2025-01-11  4:25 ` [PATCH v9 12/17] mm: move lesser used vma_area_struct members into the last cacheline Suren Baghdasaryan
2025-01-13 16:15   ` Lorenzo Stoakes
2025-01-15 10:50   ` Peter Zijlstra [this message]
2025-01-15 16:39     ` Suren Baghdasaryan
2025-02-13 22:59       ` Suren Baghdasaryan
2025-01-11  4:26 ` [PATCH v9 13/17] mm/debug: print vm_refcnt state when dumping the vma Suren Baghdasaryan
2025-01-13 16:21   ` Lorenzo Stoakes
2025-01-13 16:35     ` Liam R. Howlett
2025-01-13 17:57       ` Suren Baghdasaryan
2025-01-14 11:41         ` Lorenzo Stoakes
2025-01-11  4:26 ` [PATCH v9 14/17] mm: remove extra vma_numab_state_init() call Suren Baghdasaryan
2025-01-13 16:28   ` Lorenzo Stoakes
2025-01-13 17:56     ` Suren Baghdasaryan
2025-01-14 11:45       ` Lorenzo Stoakes
2025-01-11  4:26 ` [PATCH v9 15/17] mm: prepare lock_vma_under_rcu() for vma reuse possibility Suren Baghdasaryan
2025-01-11  4:26 ` [PATCH v9 16/17] mm: make vma cache SLAB_TYPESAFE_BY_RCU Suren Baghdasaryan
2025-01-15  2:27   ` Wei Yang
2025-01-15  3:15     ` Suren Baghdasaryan
2025-01-15  3:58       ` Liam R. Howlett
2025-01-15  5:41         ` Suren Baghdasaryan
2025-01-15  3:59       ` Mateusz Guzik
2025-01-15  5:47         ` Suren Baghdasaryan
2025-01-15  5:51           ` Mateusz Guzik
2025-01-15  6:41             ` Suren Baghdasaryan
2025-01-15  7:58       ` Vlastimil Babka
2025-01-15 15:10         ` Suren Baghdasaryan
2025-02-13 22:56           ` Suren Baghdasaryan
2025-01-15 12:17       ` Wei Yang
2025-01-15 21:46         ` Suren Baghdasaryan
2025-01-11  4:26 ` [PATCH v9 17/17] docs/mm: document latest changes to vm_lock Suren Baghdasaryan
2025-01-13 16:33   ` Lorenzo Stoakes
2025-01-13 17:56     ` Suren Baghdasaryan
2025-01-11  4:52 ` [PATCH v9 00/17] reimplement per-vma lock as a refcount Matthew Wilcox
2025-01-11  9:45   ` Suren Baghdasaryan
2025-01-13 12:14 ` Lorenzo Stoakes
2025-01-13 16:58   ` Suren Baghdasaryan
2025-01-13 17:11     ` Lorenzo Stoakes
2025-01-13 19:00       ` Suren Baghdasaryan
2025-01-14 11:35         ` Lorenzo Stoakes
2025-01-14  1:49   ` Andrew Morton
2025-01-14  2:53     ` Suren Baghdasaryan
2025-01-14  4:09       ` Andrew Morton
2025-01-14  9:09         ` Vlastimil Babka
2025-01-14 10:27           ` Hillf Danton
2025-01-14  9:47         ` Lorenzo Stoakes
2025-01-14 14:59         ` Liam R. Howlett
2025-01-14 15:54           ` Suren Baghdasaryan
2025-01-15 11:34             ` Lorenzo Stoakes
2025-01-15 15:14               ` Suren Baghdasaryan
2025-01-28  5:26 ` Shivank Garg
2025-01-28  5:50   ` Suren Baghdasaryan

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=20250115105056.GY5388@noisy.programming.kicks-ass.net \
    --to=peterz@infradead.org \
    --cc=akpm@linux-foundation.org \
    --cc=brauner@kernel.org \
    --cc=corbet@lwn.net \
    --cc=dave@stgolabs.net \
    --cc=david.laight.linux@gmail.com \
    --cc=david@redhat.com \
    --cc=dhowells@redhat.com \
    --cc=hannes@cmpxchg.org \
    --cc=hdanton@sina.com \
    --cc=hughd@google.com \
    --cc=jannh@google.com \
    --cc=kernel-team@android.com \
    --cc=klarasmodin@gmail.com \
    --cc=liam.howlett@oracle.com \
    --cc=linux-doc@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=lokeshgidra@google.com \
    --cc=lorenzo.stoakes@oracle.com \
    --cc=mgorman@techsingularity.net \
    --cc=mhocko@suse.com \
    --cc=minchan@google.com \
    --cc=mjguzik@gmail.com \
    --cc=oleg@redhat.com \
    --cc=oliver.sang@intel.com \
    --cc=pasha.tatashin@soleen.com \
    --cc=paulmck@kernel.org \
    --cc=peterx@redhat.com \
    --cc=richard.weiyang@gmail.com \
    --cc=shakeel.butt@linux.dev \
    --cc=souravpanda@google.com \
    --cc=surenb@google.com \
    --cc=vbabka@suse.cz \
    --cc=willy@infradead.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