[PATCH] mm: pmd dirty emulation in page fault handler
From: Kirill A. Shutemov <hidden>
Date: 2016-12-22 08:17:17
Also in:
linux-arch, linux-mm, stable
On Thu, Dec 22, 2016 at 08:48:21AM +0900, Minchan Kim wrote:
Andreas reported [1] made a test in jemalloc hang in THP mode in arm64.
I guess you wanted put b8d3c4c3009d before [1], right?
quoted hunk ↗ jump to hunk
http://lkml.kernel.org/r/mvmmvfy37g1.fsf at hawking.suse.de The problem is page fault handler supports only accessed flag emulation for THP page of SW-dirty/accessed architecture. This patch enables dirty-bit emulation for those architectures. Without it, MADV_FREE makes application hang by repeated fault forever. [1] mm/huge_memory.c: don't split THP page when MADV_FREE syscall is called Cc: Jason Evans <redacted> Cc: Kirill A. Shutemov <redacted> Cc: Will Deacon <redacted> Cc: Catalin Marinas <catalin.marinas@arm.com> Cc: linux-arch at vger.kernel.org Cc: linux-arm-kernel at lists.infradead.org Cc: <redacted> [4.5+] Fixes: b8d3c4c3009d ("mm/huge_memory.c: don't split THP page when MADV_FREE syscall is called") Reported-by: Andreas Schwab <redacted> Signed-off-by: Minchan Kim <minchan@kernel.org> --- mm/huge_memory.c | 6 ++++-- mm/memory.c | 18 ++++++++++-------- 2 files changed, 14 insertions(+), 10 deletions(-)diff --git a/mm/huge_memory.c b/mm/huge_memory.c index 10eedbf..29ec8a4 100644 --- a/mm/huge_memory.c +++ b/mm/huge_memory.c@@ -883,15 +883,17 @@ void huge_pmd_set_accessed(struct vm_fault *vmf, pmd_t orig_pmd) { pmd_t entry; unsigned long haddr; + bool write = vmf->flags & FAULT_FLAG_WRITE; vmf->ptl = pmd_lock(vmf->vma->vm_mm, vmf->pmd); if (unlikely(!pmd_same(*vmf->pmd, orig_pmd))) goto unlock; entry = pmd_mkyoung(orig_pmd); + if (write) + entry = pmd_mkdirty(entry); haddr = vmf->address & HPAGE_PMD_MASK; - if (pmdp_set_access_flags(vmf->vma, haddr, vmf->pmd, entry, - vmf->flags & FAULT_FLAG_WRITE)) + if (pmdp_set_access_flags(vmf->vma, haddr, vmf->pmd, entry, write)) update_mmu_cache_pmd(vmf->vma, vmf->address, vmf->pmd); unlock:diff --git a/mm/memory.c b/mm/memory.c index 36c774f..7408ddc 100644 --- a/mm/memory.c +++ b/mm/memory.c@@ -3637,18 +3637,20 @@ static int __handle_mm_fault(struct vm_area_struct *vma, unsigned long address, if (pmd_protnone(orig_pmd) && vma_is_accessible(vma)) return do_huge_pmd_numa_page(&vmf, orig_pmd); - if ((vmf.flags & FAULT_FLAG_WRITE) && - !pmd_write(orig_pmd)) { - ret = wp_huge_pmd(&vmf, orig_pmd); - if (!(ret & VM_FAULT_FALLBACK)) + if (vmf.flags & FAULT_FLAG_WRITE) { + if (!pmd_write(orig_pmd)) { + ret = wp_huge_pmd(&vmf, orig_pmd); + if (ret == VM_FAULT_FALLBACK)
In theory, more than one flag can be set and it would lead to false-negative. Bit check was the right thing. And I don't understand why do you need to change code in __handle_mm_fault() at all.
From what I see change to huge_pmd_set_accessed() should be enough.
+ goto pte_fault;
return ret;
- } else {
- huge_pmd_set_accessed(&vmf, orig_pmd);
- return 0;
+ }
}
+
+ huge_pmd_set_accessed(&vmf, orig_pmd);
+ return 0;
}
}
-
+pte_fault:
return handle_pte_fault(&vmf);
}
--
2.7.4
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo at kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email at kvack.org </a>-- Kirill A. Shutemov