Message-ID: <db6e1966-3824-45cd-8cae-740348780002@redhat.com>
Date: Tue, 19 Nov 2024 15:35:31 +0100
From: David Hildenbrand <david@...hat.com>
To: Kalesh Singh <kaleshsingh@...gle.com>
Cc: kernel-team@...roid.com, android-mm@...gle.com,
 Andrew Morton <akpm@...ux-foundation.org>, Vlastimil Babka <vbabka@...e.cz>,
 Yang Shi <yang@...amperecomputing.com>, Rik van Riel <riel@...riel.com>,
 Ryan Roberts <ryan.roberts@....com>, Suren Baghdasaryan <surenb@...gle.com>,
 Minchan Kim <minchan@...nel.org>, Hans Boehm <hboehm@...gle.com>,
 Lokesh Gidra <lokeshgidra@...gle.com>, stable@...r.kernel.org,
 "Liam R. Howlett" <Liam.Howlett@...cle.com>,
 Lorenzo Stoakes <lorenzo.stoakes@...cle.com>, Jann Horn <jannh@...gle.com>,
 Yang Shi <shy828301@...il.com>, linux-mm@...ck.org,
 linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2] mm: Respect mmap hint address when aligning for THP

On 18.11.24 22:46, Kalesh Singh wrote:
> Commit efa7df3e3bb5 ("mm: align larger anonymous mappings on THP
> boundaries") updated __get_unmapped_area() to align the start address
> for the VMA to a PMD boundary if CONFIG_TRANSPARENT_HUGEPAGE=y.
> 
> It does this by effectively looking up a region that is of size,
> request_size + PMD_SIZE, and aligning up the start to a PMD boundary.
> 
> Commit 4ef9ad19e176 ("mm: huge_memory: don't force huge page alignment
> on 32 bit") opted out of this for 32bit due to regressions in mmap base
> randomization.
> 
> Commit d4148aeab412 ("mm, mmap: limit THP alignment of anonymous
> mappings to PMD-aligned sizes") restricted this to only mmap sizes that
> are multiples of the PMD_SIZE due to reported regressions in some
> performance benchmarks -- which seemed mostly due to the reduced spatial
> locality of related mappings due to the forced PMD-alignment.
> 
> Another unintended side effect has emerged: When a user specifies an mmap
> hint address, the THP alignment logic modifies the behavior, potentially
> ignoring the hint even if a sufficiently large gap exists at the requested
> hint location.
> 
> Example Scenario:
> 
> Consider the following simplified virtual address (VA) space:
> 
>      ...
> 
>      0x200000-0x400000 --- VMA A
>      0x400000-0x600000 --- Hole
>      0x600000-0x800000 --- VMA B
> 
>      ...
> 
> A call to mmap() with hint=0x400000 and len=0x200000 behaves differently:
> 
>    - Before THP alignment: The requested region (size 0x200000) fits into
>      the gap at 0x400000, so the hint is respected.
> 
>    - After alignment: The logic searches for a region of size
>      0x400000 (len + PMD_SIZE) starting at 0x400000.
>      This search fails due to the mapping at 0x600000 (VMA B), and the hint
>      is ignored, falling back to arch_get_unmapped_area[_topdown]().
> 
> In general the hint is effectively ignored, if there is any
> existing mapping in the below range:
> 
>       [mmap_hint + mmap_size, mmap_hint + mmap_size + PMD_SIZE)
> 
> This changes the semantics of mmap hint; from ""Respect the hint if a
> sufficiently large gap exists at the requested location" to "Respect the
> hint only if an additional PMD-sized gap exists beyond the requested size".
> 
> This has performance implications for allocators that allocate their heap
> using mmap but try to keep it "as contiguous as possible" by using the
> end of the exisiting heap as the address hint. With the new behavior
> it's more likely to get a much less contiguous heap, adding extra
> fragmentation and performance overhead.
> 
> To restore the expected behavior; don't use thp_get_unmapped_area_vmflags()
> when the user provided a hint address, for anonymous mappings.
> 
> Note: As, Yang Shi, pointed out: the issue still remains for filesystems
> which are using thp_get_unmapped_area() for their get_unmapped_area() op.
> It is unclear what worklaods will regress for if we ignore THP alignment
> when the hint address is provided for such file backed mappings -- so this
> fix will be handled separately.
> 
> Cc: Andrew Morton <akpm@...ux-foundation.org>
> Cc: Vlastimil Babka <vbabka@...e.cz>
> Cc: Yang Shi <yang@...amperecomputing.com>
> Cc: Rik van Riel <riel@...riel.com>
> Cc: Ryan Roberts <ryan.roberts@....com>
> Cc: Suren Baghdasaryan <surenb@...gle.com>
> Cc: Minchan Kim <minchan@...nel.org>
> Cc: Hans Boehm <hboehm@...gle.com>
> Cc: Lokesh Gidra <lokeshgidra@...gle.com>
> Cc: <stable@...r.kernel.org>
> Fixes: efa7df3e3bb5 ("mm: align larger anonymous mappings on THP boundaries")
> Signed-off-by: Kalesh Singh <kaleshsingh@...gle.com>
> Reviewed-by: Rik van Riel <riel@...riel.com>
> Reviewed-by: Vlastimil Babka <vbabka@...e.cz>
> ---

LGTM. Hopefully that's the end of this story :)

Reviewed-by: David Hildenbrand <david@...hat.com>

-- 
Cheers,

David / dhildenb

Powered by blists - more mailing lists