linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Ryan Roberts <ryan.roberts@arm.com>
To: Andrew Morton <akpm@linux-foundation.org>,
	Anshuman Khandual <anshuman.khandual@arm.com>,
	Ard Biesheuvel <ardb@kernel.org>,
	Catalin Marinas <catalin.marinas@arm.com>,
	David Hildenbrand <david@redhat.com>,
	Greg Marsden <greg.marsden@oracle.com>,
	Ivan Ivanov <ivan.ivanov@suse.com>,
	Kalesh Singh <kaleshsingh@google.com>,
	Marc Zyngier <maz@kernel.org>,
	Mark Rutland <mark.rutland@arm.com>,
	Matthias Brugger <mbrugger@suse.com>,
	Miroslav Benes <mbenes@suse.cz>, Will Deacon <will@kernel.org>
Cc: Ryan Roberts <ryan.roberts@arm.com>,
	linux-arm-kernel@lists.infradead.org,
	linux-kernel@vger.kernel.org, linux-mm@kvack.org
Subject: [RFC PATCH v1 15/57] stackdepot: Remove PAGE_SIZE compile-time constant assumption
Date: Mon, 14 Oct 2024 11:58:22 +0100	[thread overview]
Message-ID: <20241014105912.3207374-15-ryan.roberts@arm.com> (raw)
In-Reply-To: <20241014105912.3207374-1-ryan.roberts@arm.com>

To prepare for supporting boot-time page size selection, refactor code
to remove assumptions about PAGE_SIZE being compile-time constant. Code
intended to be equivalent when compile-time page size is active.

"union handle_parts" previously calculated the number of bits required
for its pool index and offset members based on PAGE_SHIFT. This is
problematic for boot-time page size builds because the actual page size
isn't known until boot-time.

We could use PAGE_SHIFT_MAX in calculating the worst case offset bits,
but bits would be wasted that could be used for pool index when
PAGE_SIZE is set smaller than MAX, the end result being that stack depot
can address less memory than it should.

To avoid needing to dynamically define the offset and index bit widths,
let's instead fix the pool size and derive the order at runtime based on
the PAGE_SIZE. This means that the fields' widths can remain static,
with the down side being slightly increased risk of failing to allocate
the large folio.

This only affects boot-time page size builds. compile-time page size
builds will still always allocate order-2 folios.

Additionally, wrap global variables that are initialized with PAGE_SIZE
derived values using DEFINE_GLOBAL_PAGE_SIZE_VAR() so their
initialization can be deferred for boot-time page size builds.

Signed-off-by: Ryan Roberts <ryan.roberts@arm.com>
---

***NOTE***
Any confused maintainers may want to read the cover note here for context:
https://lore.kernel.org/all/20241014105514.3206191-1-ryan.roberts@arm.com/

 include/linux/stackdepot.h | 6 +++---
 lib/stackdepot.c           | 6 +++---
 2 files changed, 6 insertions(+), 6 deletions(-)

diff --git a/include/linux/stackdepot.h b/include/linux/stackdepot.h
index e9ec32fb97d4a..ac877a4e90406 100644
--- a/include/linux/stackdepot.h
+++ b/include/linux/stackdepot.h
@@ -32,10 +32,10 @@ typedef u32 depot_stack_handle_t;
 
 #define DEPOT_HANDLE_BITS (sizeof(depot_stack_handle_t) * 8)
 
-#define DEPOT_POOL_ORDER 2 /* Pool size order, 4 pages */
-#define DEPOT_POOL_SIZE (1LL << (PAGE_SHIFT + DEPOT_POOL_ORDER))
+#define DEPOT_POOL_ORDER 2 /* Pool size order, 4 pages of PAGE_SIZE_MAX */
+#define DEPOT_POOL_SIZE (1LL << (PAGE_SHIFT_MAX + DEPOT_POOL_ORDER))
 #define DEPOT_STACK_ALIGN 4
-#define DEPOT_OFFSET_BITS (DEPOT_POOL_ORDER + PAGE_SHIFT - DEPOT_STACK_ALIGN)
+#define DEPOT_OFFSET_BITS (DEPOT_POOL_ORDER + PAGE_SHIFT_MAX - DEPOT_STACK_ALIGN)
 #define DEPOT_POOL_INDEX_BITS (DEPOT_HANDLE_BITS - DEPOT_OFFSET_BITS - \
 			       STACK_DEPOT_EXTRA_BITS)
 
diff --git a/lib/stackdepot.c b/lib/stackdepot.c
index 5ed34cc963fc3..974351f0e9e3c 100644
--- a/lib/stackdepot.c
+++ b/lib/stackdepot.c
@@ -68,7 +68,7 @@ static void *new_pool;
 /* Number of pools in stack_pools. */
 static int pools_num;
 /* Offset to the unused space in the currently used pool. */
-static size_t pool_offset = DEPOT_POOL_SIZE;
+static DEFINE_GLOBAL_PAGE_SIZE_VAR(size_t, pool_offset, DEPOT_POOL_SIZE);
 /* Freelist of stack records within stack_pools. */
 static LIST_HEAD(free_stacks);
 /* The lock must be held when performing pool or freelist modifications. */
@@ -625,7 +625,7 @@ depot_stack_handle_t stack_depot_save_flags(unsigned long *entries,
 	 */
 	if (unlikely(can_alloc && !READ_ONCE(new_pool))) {
 		page = alloc_pages(gfp_nested_mask(alloc_flags),
-				   DEPOT_POOL_ORDER);
+				   get_order(DEPOT_POOL_SIZE));
 		if (page)
 			prealloc = page_address(page);
 	}
@@ -663,7 +663,7 @@ depot_stack_handle_t stack_depot_save_flags(unsigned long *entries,
 exit:
 	if (prealloc) {
 		/* Stack depot didn't use this memory, free it. */
-		free_pages((unsigned long)prealloc, DEPOT_POOL_ORDER);
+		free_pages((unsigned long)prealloc, get_order(DEPOT_POOL_SIZE));
 	}
 	if (found)
 		handle = found->handle.handle;
-- 
2.43.0



  parent reply	other threads:[~2024-10-14 11:00 UTC|newest]

Thread overview: 196+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-10-14 10:55 [RFC PATCH v1 00/57] Boot-time page size selection for arm64 Ryan Roberts
2024-10-14 10:58 ` [RFC PATCH v1 01/57] mm: Add macros ahead of supporting boot-time page size selection Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 02/57] vmlinux: Align to PAGE_SIZE_MAX Ryan Roberts
2024-10-14 16:50     ` Christoph Lameter (Ampere)
2024-10-15 10:53       ` Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 03/57] mm/memcontrol: Fix seq_buf size to save memory when PAGE_SIZE is large Ryan Roberts
2024-10-14 13:00     ` Johannes Weiner
2024-10-14 19:59     ` Shakeel Butt
2024-10-15 10:55       ` Ryan Roberts
2024-10-17 12:21         ` Michal Hocko
2024-10-17 16:09     ` Roman Gushchin
2024-10-14 10:58   ` [RFC PATCH v1 04/57] mm/page_alloc: Make page_frag_cache boot-time page size compatible Ryan Roberts
2024-11-14  8:23     ` Vlastimil Babka
2024-11-14  9:36       ` Ryan Roberts
2024-11-14  9:43         ` Vlastimil Babka
2024-10-14 10:58   ` [RFC PATCH v1 05/57] mm: Avoid split pmd ptl if pmd level is run-time folded Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 06/57] mm: Remove PAGE_SIZE compile-time constant assumption Ryan Roberts
2024-10-16 14:37     ` Ryan Roberts
2024-11-01 20:16     ` [RFC PATCH] mm/slab: Avoid build bug for calls to kmalloc with a large constant Dave Kleikamp
2024-11-06 11:44       ` Ryan Roberts
2024-11-06 15:20         ` Dave Kleikamp
2024-11-14 10:09       ` Vlastimil Babka
2024-11-26 12:18         ` Ryan Roberts
2024-11-26 12:36           ` Vlastimil Babka
2024-11-26 14:26             ` Ryan Roberts
2024-11-26 14:53             ` Ryan Roberts
2024-11-26 15:09               ` Vlastimil Babka
2024-11-26 15:27                 ` Vlastimil Babka
2024-11-26 15:33                   ` Ryan Roberts
2024-11-14 10:17     ` [RFC PATCH v1 06/57] mm: Remove PAGE_SIZE compile-time constant assumption Vlastimil Babka
2024-11-26 10:08       ` Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 07/57] fs: Introduce MAX_BUF_PER_PAGE_SIZE_MAX for array sizing Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 08/57] fs: Remove PAGE_SIZE compile-time constant assumption Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 09/57] fs/nfs: " Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 10/57] fs/ext4: " Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 11/57] fork: Permit boot-time THREAD_SIZE determination Ryan Roberts
2024-11-14 10:42     ` Vlastimil Babka
2024-10-14 10:58   ` [RFC PATCH v1 12/57] cgroup: Remove PAGE_SIZE compile-time constant assumption Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 13/57] bpf: " Ryan Roberts
2024-10-16 14:38     ` Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 14/57] pm/hibernate: " Ryan Roberts
2024-10-16 14:39     ` Ryan Roberts
2024-10-14 10:58   ` Ryan Roberts [this message]
2024-11-14 11:15     ` [RFC PATCH v1 15/57] stackdepot: " Vlastimil Babka
2024-11-26 10:15       ` Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 16/57] perf: " Ryan Roberts
2024-10-16 14:40     ` Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 17/57] kvm: " Ryan Roberts
2024-10-14 21:37     ` Sean Christopherson
2024-10-15 10:57       ` Ryan Roberts
2024-10-16 14:41     ` Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 18/57] trace: " Ryan Roberts
2024-10-14 16:46     ` Steven Rostedt
2024-10-15 11:09       ` Ryan Roberts
2024-10-18 15:24         ` Steven Rostedt
2024-10-14 10:58   ` [RFC PATCH v1 19/57] crash: " Ryan Roberts
2024-10-15  3:47     ` Baoquan He
2024-10-15 11:13       ` Ryan Roberts
2024-10-18  3:00         ` Baoquan He
2024-10-14 10:58   ` [RFC PATCH v1 20/57] crypto: " Ryan Roberts
2024-10-26  6:54     ` Herbert Xu
2024-10-14 10:58   ` [RFC PATCH v1 21/57] sunrpc: " Ryan Roberts
2024-10-16 14:42     ` Ryan Roberts
2024-10-16 14:47       ` Chuck Lever
2024-10-16 14:54         ` Jeff Layton
2024-10-16 15:09           ` Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 22/57] sound: " Ryan Roberts
2024-10-14 11:38     ` Mark Brown
2024-10-14 12:24       ` Ryan Roberts
2024-10-14 12:41         ` Takashi Iwai
2024-10-14 12:52           ` Ryan Roberts
2024-10-14 16:01         ` Mark Brown
2024-10-15 11:35           ` Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 23/57] net: " Ryan Roberts
2024-10-16 14:43     ` Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 24/57] net: fec: " Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 25/57] net: marvell: " Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 26/57] net: hns3: " Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 27/57] net: e1000: " Ryan Roberts
2024-10-16 14:43     ` Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 28/57] net: igbvf: " Ryan Roberts
2024-10-16 14:44     ` Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 29/57] net: igb: " Ryan Roberts
2024-10-16 14:45     ` Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 30/57] drivers/base: " Ryan Roberts
2024-10-16 14:45     ` Ryan Roberts
2024-10-16 15:04       ` Greg Kroah-Hartman
2024-10-16 15:12         ` Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 31/57] edac: " Ryan Roberts
2024-10-16 14:46     ` Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 32/57] optee: " Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 33/57] random: " Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 34/57] sata_sil24: " Ryan Roberts
2024-10-17  9:09     ` Niklas Cassel
2024-10-17 12:42       ` Ryan Roberts
2024-10-17 12:51         ` Niklas Cassel
2024-10-21  9:24           ` Ryan Roberts
2024-10-21 11:04             ` Niklas Cassel
2024-10-21 11:26               ` Ryan Roberts
2024-10-21 11:43                 ` Niklas Cassel
2024-10-14 10:58   ` [RFC PATCH v1 35/57] virtio: " Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 36/57] xen: " Ryan Roberts
2024-10-16 14:46     ` Ryan Roberts
2024-10-23  1:23       ` Stefano Stabellini
2024-10-24 10:32         ` Ryan Roberts
2024-10-25  1:18           ` Stefano Stabellini
2024-10-14 10:58   ` [RFC PATCH v1 37/57] arm64: Fix macros to work in C code in addition to the linker script Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 38/57] arm64: Track early pgtable allocation limit Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 39/57] arm64: Introduce macros required for boot-time page selection Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 40/57] arm64: Refactor early pgtable size calculation macros Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 41/57] arm64: Pass desired page size on command line Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 42/57] arm64: Divorce early init from PAGE_SIZE Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 43/57] arm64: Clean up simple cases of CONFIG_ARM64_*K_PAGES Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 44/57] arm64: Align sections to PAGE_SIZE_MAX Ryan Roberts
2024-10-19 14:16     ` Thomas Weißschuh
2024-10-21 11:20       ` Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 45/57] arm64: Rework trampoline rodata mapping Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 46/57] arm64: Generalize fixmap for boot-time page size Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 47/57] arm64: Statically allocate and align for worst-case " Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 48/57] arm64: Convert switch to if for non-const comparison values Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 49/57] arm64: Convert BUILD_BUG_ON to VM_BUG_ON Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 50/57] arm64: Remove PAGE_SZ asm-offset Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 51/57] arm64: Introduce cpu features for page sizes Ryan Roberts
2024-10-14 10:58   ` [RFC PATCH v1 52/57] arm64: Remove PAGE_SIZE from assembly code Ryan Roberts
2024-10-14 10:59   ` [RFC PATCH v1 53/57] arm64: Runtime-fold pmd level Ryan Roberts
2024-10-14 10:59   ` [RFC PATCH v1 54/57] arm64: Support runtime folding in idmap_kpti_install_ng_mappings Ryan Roberts
2024-10-14 10:59   ` [RFC PATCH v1 55/57] arm64: TRAMP_VALIAS is no longer compile-time constant Ryan Roberts
2024-10-14 11:21     ` Ard Biesheuvel
2024-10-14 11:28       ` Ryan Roberts
2024-10-14 10:59   ` [RFC PATCH v1 56/57] arm64: Determine THREAD_SIZE at boot-time Ryan Roberts
2024-10-14 10:59   ` [RFC PATCH v1 57/57] arm64: Enable boot-time page size selection Ryan Roberts
2024-10-15 17:42     ` Zi Yan
2024-10-16  8:14       ` Ryan Roberts
2024-10-16 14:21         ` Zi Yan
2024-10-16 14:31           ` Ryan Roberts
2024-10-16 14:35             ` Zi Yan
2024-10-15 17:52     ` Michael Kelley
2024-10-16  8:17       ` Ryan Roberts
2024-10-14 13:54   ` [RFC PATCH v1 01/57] mm: Add macros ahead of supporting " Pingfan Liu
2024-10-14 14:07     ` Ryan Roberts
2024-10-15  3:04       ` Pingfan Liu
2024-10-15 11:16         ` Ryan Roberts
2024-10-16 14:36   ` Ryan Roberts
2024-10-30  8:45   ` Ryan Roberts
2024-10-14 17:32 ` [RFC PATCH v1 00/57] Boot-time page size selection for arm64 Florian Fainelli
2024-10-15 11:48   ` Ryan Roberts
2024-10-15 18:38 ` Michael Kelley
2024-10-16  8:23   ` Ryan Roberts
2024-10-16 15:16 ` David Hildenbrand
2024-10-16 16:08   ` Ryan Roberts
2024-10-17 12:27 ` Petr Tesarik
2024-10-17 12:32   ` Ryan Roberts
2024-10-18 12:56     ` Petr Tesarik
2024-10-18 14:41       ` Petr Tesarik
2024-10-21 11:47         ` Ryan Roberts
2024-10-23 21:00     ` Thomas Tai
2024-10-24 10:48       ` Ryan Roberts
2024-10-24 11:45         ` Petr Tesarik
2024-10-24 12:10           ` Ryan Roberts
2024-10-30 22:11         ` Sumit Gupta
2024-11-11 12:14     ` Petr Tesarik
2024-11-11 12:25       ` Ryan Roberts
2024-11-12  9:45         ` Petr Tesarik
2024-11-12 10:19           ` Ryan Roberts
2024-11-12 10:50             ` Petr Tesarik
2024-11-13 12:40               ` Petr Tesarik
2024-11-13 12:56                 ` Ryan Roberts
2024-11-13 14:22                   ` Petr Tesarik
2024-12-05 17:20     ` Petr Tesarik
2024-12-05 18:52       ` Michael Kelley
2024-12-06  7:50         ` Petr Tesarik
2024-12-06 10:26           ` Ryan Roberts
2024-12-06 13:05             ` Michael Kelley
2024-10-17 22:05 ` Dave Kleikamp
2024-10-21 11:49   ` Ryan Roberts
2024-10-18 18:15 ` Joseph Salisbury
2024-10-18 18:27   ` David Hildenbrand
2024-10-18 19:19     ` [External] : " Joseph Salisbury
2024-10-18 19:27       ` David Hildenbrand
2024-10-18 20:06         ` Joseph Salisbury
2024-10-21  9:55           ` Ryan Roberts
2024-10-19 15:47 ` Neal Gompa
2024-10-21 11:02   ` Ryan Roberts
2024-10-21 11:32     ` Eric Curtin
2024-10-21 11:51       ` Ryan Roberts
2024-10-21 13:49         ` Neal Gompa
2024-10-21 15:01           ` Ryan Roberts
2024-10-22  9:33             ` Neal Gompa
2024-10-22 15:03               ` Nick Chan
2024-10-22 15:12                 ` Ryan Roberts
2024-10-22 17:30                   ` Neal Gompa
2024-10-24 10:34                     ` Ryan Roberts
2024-10-31 21:07 ` Catalin Marinas
2024-11-06 11:37   ` Ryan Roberts
2024-11-07 12:35     ` Catalin Marinas
2024-11-07 12:47       ` Ryan Roberts

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=20241014105912.3207374-15-ryan.roberts@arm.com \
    --to=ryan.roberts@arm.com \
    --cc=akpm@linux-foundation.org \
    --cc=anshuman.khandual@arm.com \
    --cc=ardb@kernel.org \
    --cc=catalin.marinas@arm.com \
    --cc=david@redhat.com \
    --cc=greg.marsden@oracle.com \
    --cc=ivan.ivanov@suse.com \
    --cc=kaleshsingh@google.com \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=mark.rutland@arm.com \
    --cc=maz@kernel.org \
    --cc=mbenes@suse.cz \
    --cc=mbrugger@suse.com \
    --cc=will@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