linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Wupeng Ma <mawupeng1@huawei.com>
To: <catalin.marinas@arm.com>, <will@kernel.org>,
	<ryan.roberts@arm.com>, <anshuman.khandual@arm.com>,
	<akpm@linux-foundation.org>
Cc: <peterx@redhat.com>, <joey.gouly@arm.com>,
	<yangyicong@hisilicon.com>, <ioworker0@gmail.com>,
	<baohua@kernel.org>, <jack@suse.cz>,
	<linux-arm-kernel@lists.infradead.org>, <linux-mm@kvack.org>,
	<linux-kernel@vger.kernel.org>, Wupeng Ma <mawupeng1@huawei.com>
Subject: [PATCH] arm64: dax: add devmap check for pmd_trans_huge
Date: Tue, 8 Apr 2025 16:59:14 +0800	[thread overview]
Message-ID: <20250408085914.1946183-1-mawupeng1@huawei.com> (raw)

During our test in ext4 dax linux-v5.10 on arm64. A BUG_ON is trigger in
follow_invalidate_pte as follow since this pmd is seem as pmd_trans_huge.
However this page is really a dax-pmds rather than a pmd trans huge.

Call trace is shown as follow:

  ------------[ cut here ]------------
  kernel BUG at mm/memory.c:5185!
  CPU: 0 PID: 150 Comm: kworker/u8:10 Not tainted 5.10.0-01678-g1e62aad66bbc-dirty #36
  pc : follow_invalidate_pte+0xdc/0x5e0
  lr : follow_invalidate_pte+0xc4/0x5e0
  sp : ffffa00012997110
  Call trace:
   follow_invalidate_pte+0xdc/0x5e0
   dax_entry_mkclean+0x250/0x870
   dax_writeback_one+0xac/0x380
   dax_writeback_mapping_range+0x22c/0x704
   ext4_dax_writepages+0x234/0x6e4
   do_writepages+0xc8/0x1c0
   __writeback_single_inode+0xb8/0x560
   writeback_sb_inodes+0x344/0x7a0
   wb_writeback+0x1f8/0x6b0
   wb_do_writeback+0x194/0x3cc
   wb_workfn+0x14c/0x590
   process_one_work+0x470/0xa30
   worker_thread+0xac/0x510
   kthread+0x1e0/0x220
   ret_from_fork+0x10/0x18
  ---[ end trace 0f479050bd4b1818 ]---
  Kernel panic - not syncing: Oops - BUG: Fatal exception
  ---[ end Kernel panic - not syncing: Oops - BUG: Fatal exception ]---

Commit 5c7fb56e5e3f ("mm, dax: dax-pmd vs thp-pmd vs hugetlbfs-pmd") and
commit 36b78402d97a ("powerpc/hash64/devmap: Use H_PAGE_THP_HUGE when
setting up huge devmap PTE entries") already check pmd_devmap during
checking pmd_trans_huge. Since pmd_devmap() is used to distinguish dax-pmds,
add the same check for arm64 to fix this problem.

Add PTE_DEVMAP in pte_modify as commit 4628a64591e6 ("mm: Preserve
_PAGE_DEVMAP across mprotect() calls") does to avoid the same issue in
mprotect.

Fixes: 73b20c84d42d ("arm64: mm: implement pte_devmap support")
Signed-off-by: Wupeng Ma <mawupeng1@huawei.com>
---
 arch/arm64/include/asm/pgtable.h | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/arch/arm64/include/asm/pgtable.h b/arch/arm64/include/asm/pgtable.h
index d3b538be1500b..b9a618127c01b 100644
--- a/arch/arm64/include/asm/pgtable.h
+++ b/arch/arm64/include/asm/pgtable.h
@@ -740,7 +740,7 @@ static inline int pmd_trans_huge(pmd_t pmd)
 	 * as a table, so force the valid bit for the comparison.
 	 */
 	return pmd_val(pmd) && pmd_present(pmd) &&
-	       !pmd_table(__pmd(pmd_val(pmd) | PTE_VALID));
+	       !pmd_table(__pmd(pmd_val(pmd) | PTE_VALID)) && !pmd_devmap(pmd);
 }
 #endif /* CONFIG_TRANSPARENT_HUGEPAGE */
 
@@ -1186,7 +1186,8 @@ static inline pte_t pte_modify(pte_t pte, pgprot_t newprot)
 	 */
 	const pteval_t mask = PTE_USER | PTE_PXN | PTE_UXN | PTE_RDONLY |
 			      PTE_PRESENT_INVALID | PTE_VALID | PTE_WRITE |
-			      PTE_GP | PTE_ATTRINDX_MASK | PTE_PO_IDX_MASK;
+			      PTE_GP | PTE_ATTRINDX_MASK | PTE_PO_IDX_MASK |
+			      PTE_DEVMAP;
 
 	/* preserve the hardware dirty information */
 	if (pte_hw_dirty(pte))
-- 
2.43.0



             reply	other threads:[~2025-04-08  9:09 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-04-08  8:59 Wupeng Ma [this message]
2025-04-08 23:05 ` Alistair Popple
2025-04-09  1:08   ` mawupeng

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=20250408085914.1946183-1-mawupeng1@huawei.com \
    --to=mawupeng1@huawei.com \
    --cc=akpm@linux-foundation.org \
    --cc=anshuman.khandual@arm.com \
    --cc=baohua@kernel.org \
    --cc=catalin.marinas@arm.com \
    --cc=ioworker0@gmail.com \
    --cc=jack@suse.cz \
    --cc=joey.gouly@arm.com \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=peterx@redhat.com \
    --cc=ryan.roberts@arm.com \
    --cc=will@kernel.org \
    --cc=yangyicong@hisilicon.com \
    /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