From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mail-qt1-f182.google.com (mail-qt1-f182.google.com [209.85.160.182]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 7B9EC158520 for ; Tue, 25 Feb 2025 17:02:49 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.160.182 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1740502971; cv=none; b=iVgkjvrE9xFBy21AOIOvEzo/YUg+GJ4FZA1QbQ5S3pjaMjq9eoiwYTBvRL0AMnn+7mknrcn20YdrKSalSsTgfBMPVk8SXt7T6MygAHDoIXSA1S68LCNrToEsh3IcTWVYathzMz6+BwZn3IgCzdYmYWLAVROU0yYrCkId94o1ap8= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1740502971; c=relaxed/simple; bh=SkFG7jbh4sZ1Hhr/lirAJj6lsLFj3YfkDYM6mN2hmTQ=; h=MIME-Version:References:In-Reply-To:From:Date:Message-ID:Subject: To:Cc:Content-Type; b=tX8buzsHi62i8BQroSNc90tRNYD2pP/rLNS9v4a2DjOf5S0LaSlKNM5DuJpVdCEu7guuAd+5OFf1ucwa3P65xa3qOzuvX7R0y4YDxnrNjnvWDb86qo0hMmNIwhAY7sQMoqDFmS+f2VPnKcmwceQwgn4ycv0o6El357DLrooFR8Q= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com; spf=pass smtp.mailfrom=google.com; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b=LXb0T+AZ; arc=none smtp.client-ip=209.85.160.182 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=google.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b="LXb0T+AZ" Received: by mail-qt1-f182.google.com with SMTP id d75a77b69052e-472098e6e75so374981cf.1 for ; Tue, 25 Feb 2025 09:02:49 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20230601; t=1740502968; x=1741107768; darn=vger.kernel.org; h=content-transfer-encoding:cc:to:subject:message-id:date:from :in-reply-to:references:mime-version:from:to:cc:subject:date :message-id:reply-to; bh=fFlvzy5MlROQrAEfjkrWxFOVsVqFB4jhR5ycbT0zPr8=; b=LXb0T+AZ1wyjK+oejo7/8qJOPWvo9PqdFRk4T1mslN0ICH/qC9z9zuQN2M7ZSRm+9t +N3AAZQcDrksBVVoBPFLGxrFJNcgoNx2nq9mgEIOP32CLr5ke2C7Z3kn4NT3LofFHXcN awUy3xG8xBhpX4TehdcWtREgxBA1GF4TOc+Wkbby2zySQZn1T4A1qTK1oip1Kk2VREcQ Ff200zKDfQBLYT5huBVwpqQbVvNBMOm8PyRcFOjdS1C60ALhmOqzIRYTJvfIyexKXAtW Eg/moypUpGeufYcquW6K3XOqTZTdgyP6DfVmEwUMzzmQ5JJzXm5CSjbFQrLdQmAQjZOs RfPA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1740502968; x=1741107768; h=content-transfer-encoding:cc:to:subject:message-id:date:from :in-reply-to:references:mime-version:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=fFlvzy5MlROQrAEfjkrWxFOVsVqFB4jhR5ycbT0zPr8=; b=H7wd8CMH266FdIPNcnDx/21F7KxQacL/C6JO95AApONlh4Pvod8dpa1PiM2GHII1IV JjdIaiLpe/l8TUYoLxWMYGIzm7wsaf3bdhMPMeSNkxvRtH4HSsn+yEnkfoWGA/KR2O7c c088+TvfnUrZO2wHsKB2b3L8oLOvpXVeuQjBSZrszCo4IvFV7GobeL/7vzuwhO+AmWVS G89/jL+bBh5aUUkdCZIj7U9yXV3YJs+IkiECRH7FYxepvufKZ33XCrjmCsgThubRAwM2 UaZN8W+N5Jm+cCuG+NmVp+sleEvtEKsbTmsxXMbLJ7jtWK4lm1a5+1bDYT4LkHESX8qg Z+PA== X-Forwarded-Encrypted: i=1; AJvYcCWn2l8vi7FZatWRABpVyLN2a1xV4A27Y8Dp2HGs/Qlz/zEXAuJqPAOmZEReCRkAC4HMNhUFocCvKDyM5Fg=@vger.kernel.org X-Gm-Message-State: AOJu0YyXwH3QaRSaDzDl9YrQ4qNhjy7hHsKmQRitDUBODGFWifduJ+dA G+tIjwduC+QK61Y/f1+ZKU4JK4GUL6pgnitq3JZw2zv2r2O2uBJBPBuoMQvsQ43CQRqSqM+AIrs 1f6UNN4Tt7SDXv7j+OTl7A2PAvOinfEDUucic X-Gm-Gg: ASbGncvQmsJNCPS2yJ+B/m/kdsLTH+9/lcef5WtB6vbQyO0aEYxZma6WoDJPv9XcfGj EG1PotLlNQPsr040cTFU7SeUxMzcGvdvtxw5HTy0GZPcwmwK+zpB6fZbz3Y5cZpNR/IbHHV11OG VVcuKmsUA= X-Google-Smtp-Source: AGHT+IGfZoydOPbzChKno0IgijhTP/lntU/5d7TQ/cYmx5FO/mGg9Yt2l61INuKodZ8dIVTTLVvYi51ees46MJMOkYA= X-Received: by 2002:a05:622a:24f:b0:472:796:9ea6 with SMTP id d75a77b69052e-47376e8faf1mr4697511cf.17.1740502967850; Tue, 25 Feb 2025 09:02:47 -0800 (PST) Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 References: <20250225110525.26732-1-21cnbao@gmail.com> In-Reply-To: From: Suren Baghdasaryan Date: Tue, 25 Feb 2025 09:02:35 -0800 X-Gm-Features: AWEUYZkV01cVQhPz2VcatiMKz9-_P9bQisDfhrwc_BjMe77z5mBkffDUdqe8knA Message-ID: Subject: Re: [PATCH RFC] mm: Fix kernel BUG when userfaultfd_move encounters swapcache To: Peter Xu Cc: Barry Song <21cnbao@gmail.com>, Liam.Howlett@oracle.com, aarcange@redhat.com, akpm@linux-foundation.org, axelrasmussen@google.com, bgeffon@google.com, brauner@kernel.org, david@redhat.com, hughd@google.com, jannh@google.com, kaleshsingh@google.com, linux-kernel@vger.kernel.org, linux-mm@kvack.org, lokeshgidra@google.com, mhocko@suse.com, ngeoffray@google.com, rppt@kernel.org, ryan.roberts@arm.com, shuah@kernel.org, v-songbaohua@oppo.com, viro@zeniv.linux.org.uk, willy@infradead.org, yuzhao@google.com, zhangpeng362@huawei.com, zhengtangquan@oppo.com Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable On Tue, Feb 25, 2025 at 7:34=E2=80=AFAM Peter Xu wrote: > > On Wed, Feb 26, 2025 at 12:05:25AM +1300, Barry Song wrote: > > On Fri, Feb 21, 2025 at 1:36=E2=80=AFPM Suren Baghdasaryan wrote: > > > > > > On Thu, Feb 20, 2025 at 3:52=E2=80=AFPM Suren Baghdasaryan wrote: > > > > > > > > On Thu, Feb 20, 2025 at 3:47=E2=80=AFPM Suren Baghdasaryan wrote: > > > > > > > > > > On Thu, Feb 20, 2025 at 2:59=E2=80=AFPM Peter Xu wrote: > > > > > > > > > > > > On Thu, Feb 20, 2025 at 12:04:40PM +1300, Barry Song wrote: > > > > > > > On Thu, Feb 20, 2025 at 11:15=E2=80=AFAM Peter Xu wrote: > > > > > > > > > > > > > > > > On Thu, Feb 20, 2025 at 09:37:50AM +1300, Barry Song wrote: > > > > > > > > > On Thu, Feb 20, 2025 at 7:27=E2=80=AFAM Suren Baghdasarya= n wrote: > > > > > > > > > > > > > > > > > > > > On Wed, Feb 19, 2025 at 3:25=E2=80=AFAM Barry Song <21c= nbao@gmail.com> wrote: > > > > > > > > > > > > > > > > > > > > > > From: Barry Song > > > > > > > > > > > > > > > > > > > > > > userfaultfd_move() checks whether the PTE entry is pr= esent or a > > > > > > > > > > > swap entry. > > > > > > > > > > > > > > > > > > > > > > - If the PTE entry is present, move_present_pte() han= dles folio > > > > > > > > > > > migration by setting: > > > > > > > > > > > > > > > > > > > > > > src_folio->index =3D linear_page_index(dst_vma, dst= _addr); > > > > > > > > > > > > > > > > > > > > > > - If the PTE entry is a swap entry, move_swap_pte() s= imply copies > > > > > > > > > > > the PTE to the new dst_addr. > > > > > > > > > > > > > > > > > > > > > > This approach is incorrect because even if the PTE is= a swap > > > > > > > > > > > entry, it can still reference a folio that remains in= the swap > > > > > > > > > > > cache. > > > > > > > > > > > > > > > > > > > > > > If do_swap_page() is triggered, it may locate the fol= io in the > > > > > > > > > > > swap cache. However, during add_rmap operations, a ke= rnel panic > > > > > > > > > > > can occur due to: > > > > > > > > > > > page_pgoff(folio, page) !=3D linear_page_index(vma, = address) > > > > > > > > > > > > > > > > > > > > Thanks for the report and reproducer! > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > $./a.out > /dev/null > > > > > > > > > > > [ 13.336953] page: refcount:6 mapcount:1 mapping:00= 000000f43db19c index:0xffffaf150 pfn:0x4667c > > > > > > > > > > > [ 13.337520] head: order:2 mapcount:1 entire_mapcou= nt:0 nr_pages_mapped:1 pincount:0 > > > > > > > > > > > [ 13.337716] memcg:ffff00000405f000 > > > > > > > > > > > [ 13.337849] anon flags: 0x3fffc0000020459(locked|u= ptodate|dirty|owner_priv_1|head|swapbacked|node=3D0|zone=3D0|lastcpupid=3D0= xffff) > > > > > > > > > > > [ 13.338630] raw: 03fffc0000020459 ffff80008507b538= ffff80008507b538 ffff000006260361 > > > > > > > > > > > [ 13.338831] raw: 0000000ffffaf150 0000000000004000= 0000000600000000 ffff00000405f000 > > > > > > > > > > > [ 13.339031] head: 03fffc0000020459 ffff80008507b53= 8 ffff80008507b538 ffff000006260361 > > > > > > > > > > > [ 13.339204] head: 0000000ffffaf150 000000000000400= 0 0000000600000000 ffff00000405f000 > > > > > > > > > > > [ 13.339375] head: 03fffc0000000202 fffffdffc0199f0= 1 ffffffff00000000 0000000000000001 > > > > > > > > > > > [ 13.339546] head: 0000000000000004 000000000000000= 0 00000000ffffffff 0000000000000000 > > > > > > > > > > > [ 13.339736] page dumped because: VM_BUG_ON_PAGE(pa= ge_pgoff(folio, page) !=3D linear_page_index(vma, address)) > > > > > > > > > > > [ 13.340190] ------------[ cut here ]------------ > > > > > > > > > > > [ 13.340316] kernel BUG at mm/rmap.c:1380! > > > > > > > > > > > [ 13.340683] Internal error: Oops - BUG: 00000000f2= 000800 [#1] PREEMPT SMP > > > > > > > > > > > [ 13.340969] Modules linked in: > > > > > > > > > > > [ 13.341257] CPU: 1 UID: 0 PID: 107 Comm: a.out Not= tainted 6.14.0-rc3-gcf42737e247a-dirty #299 > > > > > > > > > > > [ 13.341470] Hardware name: linux,dummy-virt (DT) > > > > > > > > > > > [ 13.341671] pstate: 60000005 (nZCv daif -PAN -UAO = -TCO -DIT -SSBS BTYPE=3D--) > > > > > > > > > > > [ 13.341815] pc : __page_check_anon_rmap+0xa0/0xb0 > > > > > > > > > > > [ 13.341920] lr : __page_check_anon_rmap+0xa0/0xb0 > > > > > > > > > > > [ 13.342018] sp : ffff80008752bb20 > > > > > > > > > > > [ 13.342093] x29: ffff80008752bb20 x28: fffffdffc01= 99f00 x27: 0000000000000001 > > > > > > > > > > > [ 13.342404] x26: 0000000000000000 x25: 00000000000= 00001 x24: 0000000000000001 > > > > > > > > > > > [ 13.342575] x23: 0000ffffaf0d0000 x22: 0000ffffaf0= d0000 x21: fffffdffc0199f00 > > > > > > > > > > > [ 13.342731] x20: fffffdffc0199f00 x19: ffff0000062= 10700 x18: 00000000ffffffff > > > > > > > > > > > [ 13.342881] x17: 6c203d2120296567 x16: 6170202c6f6= 96c6f x15: 662866666f67705f > > > > > > > > > > > [ 13.343033] x14: 6567617028454741 x13: 29297373657= 26464 x12: ffff800083728ab0 > > > > > > > > > > > [ 13.343183] x11: ffff800082996bf8 x10: 00000000000= 00fd7 x9 : ffff80008011bc40 > > > > > > > > > > > [ 13.343351] x8 : 0000000000017fe8 x7 : 00000000fff= ff000 x6 : ffff8000829eebf8 > > > > > > > > > > > [ 13.343498] x5 : c0000000fffff000 x4 : 00000000000= 00000 x3 : 0000000000000000 > > > > > > > > > > > [ 13.343645] x2 : 0000000000000000 x1 : ffff0000062= db980 x0 : 000000000000005f > > > > > > > > > > > [ 13.343876] Call trace: > > > > > > > > > > > [ 13.344045] __page_check_anon_rmap+0xa0/0xb0 (P) > > > > > > > > > > > [ 13.344234] folio_add_anon_rmap_ptes+0x22c/0x320 > > > > > > > > > > > [ 13.344333] do_swap_page+0x1060/0x1400 > > > > > > > > > > > [ 13.344417] __handle_mm_fault+0x61c/0xbc8 > > > > > > > > > > > [ 13.344504] handle_mm_fault+0xd8/0x2e8 > > > > > > > > > > > [ 13.344586] do_page_fault+0x20c/0x770 > > > > > > > > > > > [ 13.344673] do_translation_fault+0xb4/0xf0 > > > > > > > > > > > [ 13.344759] do_mem_abort+0x48/0xa0 > > > > > > > > > > > [ 13.344842] el0_da+0x58/0x130 > > > > > > > > > > > [ 13.344914] el0t_64_sync_handler+0xc4/0x138 > > > > > > > > > > > [ 13.345002] el0t_64_sync+0x1ac/0x1b0 > > > > > > > > > > > [ 13.345208] Code: aa1503e0 f000f801 910f6021 97ff5= 779 (d4210000) > > > > > > > > > > > [ 13.345504] ---[ end trace 0000000000000000 ]--- > > > > > > > > > > > [ 13.345715] note: a.out[107] exited with irqs disa= bled > > > > > > > > > > > [ 13.345954] note: a.out[107] exited with preempt_c= ount 2 > > > > > > > > > > > > > > > > > > > > > > Fully fixing it would be quite complex, requiring sim= ilar handling > > > > > > > > > > > of folios as done in move_present_pte. > > > > > > > > > > > > > > > > > > > > How complex would that be? Is it a matter of adding > > > > > > > > > > folio_maybe_dma_pinned() checks, doing folio_move_anon_= rmap() and > > > > > > > > > > folio->index =3D linear_page_index like in move_present= _pte() or > > > > > > > > > > something more? > > > > > > > > > > > > > > > > > > My main concern is still with large folios that require a= split_folio() > > > > > > > > > during move_pages(), as the entire folio shares the same = index and > > > > > > > > > anon_vma. However, userfaultfd_move() moves pages individ= ually, > > > > > > > > > making a split necessary. > > > > > > > > > > > > > > > > > > However, in split_huge_page_to_list_to_order(), there is = a: > > > > > > > > > > > > > > > > > > if (folio_test_writeback(folio)) > > > > > > > > > return -EBUSY; > > > > > > > > > > > > > > > > > > This is likely true for swapcache, right? However, even f= or move_present_pte(), > > > > > > > > > it simply returns -EBUSY: > > > > > > > > > > > > > > > > > > move_pages_pte() > > > > > > > > > { > > > > > > > > > /* at this point we have src_folio locked= */ > > > > > > > > > if (folio_test_large(src_folio)) { > > > > > > > > > /* split_folio() can block */ > > > > > > > > > pte_unmap(&orig_src_pte); > > > > > > > > > pte_unmap(&orig_dst_pte); > > > > > > > > > src_pte =3D dst_pte =3D NULL; > > > > > > > > > err =3D split_folio(src_folio); > > > > > > > > > if (err) > > > > > > > > > goto out; > > > > > > > > > > > > > > > > > > /* have to reacquire the folio af= ter it got split */ > > > > > > > > > folio_unlock(src_folio); > > > > > > > > > folio_put(src_folio); > > > > > > > > > src_folio =3D NULL; > > > > > > > > > goto retry; > > > > > > > > > } > > > > > > > > > } > > > > > > > > > > > > > > > > > > Do we need a folio_wait_writeback() before calling split_= folio()? > > > > > > > > > > > > > > > > Maybe no need in the first version to fix the immediate bug= ? > > > > > > > > > > > > > > > > It's also not always the case to hit writeback here. IIUC, = writeback only > > > > > > > > happens for a short window when the folio was just added in= to swapcache. > > > > > > > > MOVE can happen much later after that anytime before a swap= in. My > > > > > > > > understanding is that's also what Matthew wanted to point o= ut. It may be > > > > > > > > better justified of that in a separate change with some per= formance > > > > > > > > measurements. > > > > > > > > > > > > > > The bug we=E2=80=99re discussing occurs precisely within the = short window you > > > > > > > mentioned. > > > > > > > > > > > > > > 1. add_to_swap: The folio is added to swapcache. > > > > > > > 2. try_to_unmap: PTEs are converted to swap entries. > > > > > > > 3. pageout > > > > > > > 4. Swapcache is cleared. > > > > > > > > > > > > Hmm, I see. I was expecting step 4 to be "writeback is cleared"= .. or at > > > > > > least that should be step 3.5, as IIUC "writeback" needs to be = cleared > > > > > > before "swapcache" bit being cleared. > > > > > > > > > > > > > > > > > > > > The issue happens between steps 2 and 4, where the PTE is not= present, but > > > > > > > the folio is still in swapcache - the current code does move_= swap_pte() but does > > > > > > > not fixup folio->index within swapcache. > > > > > > > > > > > > One thing I'm still not clear here is why it's a race condition= , rather > > > > > > than more severe than that. I mean, folio->index is definitely= wrong, then > > > > > > as long as the page still in swapcache, we should be able to mo= ve the swp > > > > > > entry over to dest addr of UFFDIO_MOVE, read on dest addr, then= it'll see > > > > > > the page in swapcache with the wrong folio->index already and t= rigger. > > > > > > > > > > > > I wrote a quick test like that, it actually won't trigger.. > > > > > > > > > > > > I had a closer look in the code, I think it's because do_swap_p= age() has > > > > > > the logic to detect folio->index matching first, and allocate a= new folio > > > > > > if it doesn't match in ksm_might_need_to_copy(). IIUC that was= for > > > > > > ksm.. but it looks like it's functioning too here. > > > > > > > > > > > > ksm_might_need_to_copy: > > > > > > if (folio_test_ksm(folio)) { > > > > > > if (folio_stable_node(folio) && > > > > > > !(ksm_run & KSM_RUN_UNMERGE)) > > > > > > return folio; /* no need to copy it *= / > > > > > > } else if (!anon_vma) { > > > > > > return folio; /* no need to copy it *= / > > > > > > } else if (folio->index =3D=3D linear_page_index(vma, a= ddr) && <---------- [1] > > > > > > anon_vma->root =3D=3D vma->anon_vma->ro= ot) { > > > > > > return folio; /* still no need to cop= y it */ > > > > > > } > > > > > > ... > > > > > > > > > > > > new_folio =3D vma_alloc_folio(GFP_HIGHUSER_MOVABLE, 0, = vma, addr); <---- [2] > > > > > > ... > > > > > > > > > > > > So I believe what I hit is at [1] it sees index doesn't match, = then it > > > > > > decided to allocate a new folio. In this case, it won't hit yo= ur BUG > > > > > > because it'll be "folio !=3D swapcache" later, so it'll setup t= he > > > > > > folio->index for the new one, rather than the sanity check. > > > > > > > > > > > > Do you know how your case got triggered, being able to bypass t= he above [1] > > > > > > which should check folio->index already? > > > > > > > > > > To understand the change I tried applying the proposed patch to b= oth > > > > > mm-unstable and Linus' ToT and got conflicts for both trees. Barr= y, > > > > > which baseline are you using? > > > > > > > > Oops, never mind. My mistake. Copying from the email messed up tabs= ... > > > > It applies cleanly. > > > > > > Overall the code seems correct to me, however the new code has quite > > > complex logical structure IMO. Original simplified code structure is > > > like this: > > > > > > if (pte_present(orig_src_pte)) { > > > if (is_zero_pfn) { > > > move_zeropage_pte() > > > return > > > } > > > // pin and lock src_folio > > > spin_lock(src_ptl) > > > folio_get(folio) > > > folio_trylock(folio) > > > if (folio_test_large(src_folio)) > > > split_folio(src_folio) > > > anon_vma_trylock_write(src_anon_vma) > > > move_present_pte() > > > } else { > > > if (non_swap_entry(entry)) > > > if (is_migration_entry(entry)) > > > handle migration entry > > > else > > > move_swap_pte() > > > } > > > > > > The new structure looks like this: > > > > > > if (!pte_present(orig_src_pte)) { > > > if (is_migration_entry(entry)) { > > > handle migration entry > > > return > > > } > > > if (!non_swap_entry() || !pte_swp_exclusive()) > > > return > > > si =3D get_swap_device(entry); > > > } > > > if (pte_present(orig_src_pte) && is_zero_pfn(pte_pfn(orig_src_pte))) > > > move_zeropage_pte() > > > return > > > } > > > pin and lock src_folio > > > spin_lock(src_ptl) > > > if (pte_present(orig_src_pte)) > > > folio_get(folio) > > > else { > > > folio =3D filemap_get_folio(swap_entry) > > > if (IS_ERR(folio)) > > > move_swap_pte() > > > return > > > } > > > } > > > folio_trylock(folio) > > > if (folio_test_large(src_folio)) > > > split_folio(src_folio) > > > if (pte_present(orig_src_pte)) > > > anon_vma_trylock_write(src_anon_vma) > > > move_pte_and_folio() > > > > > > This looks more complex and harder to follow. Might be the reason > > > David was not in favour of treating swapcache and present pages in th= e > > > same path. And now I would agree that refactoring some common parts > > > and not breaking the original structure might be cleaner. > > > > Exactly, that=E2=80=99s the cost we=E2=80=99re facing in trying to shar= e the code path > > for swap and present PTEs. > > > > I tried to extract some common functions for present PTE and swap entri= es, > > but I found too many detailed differences and variants. This made the c= ommon > > function overly complex, turning it into a real "monster." As a result,= I > > don't think this approach would make the code any more readable or clea= ner. > > > > After trying a couple of times, I feel the following is somehow more > > readable: > > (Lokesh is eager for the small folios fixes to be merged without furthe= r > > delay. So, I'd prefer to return -EBUSY for large folios in the hotfixes > > and handle the mTHP -EBUSY issue in a separate patch later.) Hi Barry, This is much more readable, thank you! With Peter's comments addressed I think this is ready to be posted as an official fix. Thanks, Suren. > > > > diff --git a/mm/userfaultfd.c b/mm/userfaultfd.c > > index 867898c4e30b..eed9286ec1f3 100644 > > --- a/mm/userfaultfd.c > > +++ b/mm/userfaultfd.c > > @@ -18,6 +18,7 @@ > > #include > > #include > > #include "internal.h" > > +#include "swap.h" > > > > static __always_inline > > bool validate_dst_vma(struct vm_area_struct *dst_vma, unsigned long ds= t_end) > > @@ -1072,15 +1073,15 @@ static int move_present_pte(struct mm_struct *m= m, > > return err; > > } > > > > -static int move_swap_pte(struct mm_struct *mm, > > +static int move_swap_pte(struct mm_struct *mm, struct vm_area_struct *= dst_vma, > > unsigned long dst_addr, unsigned long src_addr, > > pte_t *dst_pte, pte_t *src_pte, > > pte_t orig_dst_pte, pte_t orig_src_pte, > > pmd_t *dst_pmd, pmd_t dst_pmdval, > > - spinlock_t *dst_ptl, spinlock_t *src_ptl) > > + spinlock_t *dst_ptl, spinlock_t *src_ptl, > > + struct folio *src_folio) > > { > > - if (!pte_swp_exclusive(orig_src_pte)) > > - return -EBUSY; > > + int err =3D 0; > > > > double_pt_lock(dst_ptl, src_ptl); > > > > @@ -1090,11 +1091,22 @@ static int move_swap_pte(struct mm_struct *mm, > > return -EAGAIN; > > } > > > > + if (src_folio) { > > We'd better add a comment here explaining src_folio in this case is a swa= p > cache folio, and we're updating index to make sure if the folio can be > reused later in a swapin, the rmap info will match, or something like tha= t. > > > + /* Folio got pinned from under us. Put it back and fail t= he move. */ > > + if (folio_maybe_dma_pinned(src_folio)) { > > + err =3D -EBUSY; > > + goto out; > > + } > > If the swap entry is guaranteed exclusive (which I think will hold true),= I > think we can drop this and the "out" label, as it can't happen. > > > + folio_move_anon_rmap(src_folio, dst_vma); > > + src_folio->index =3D linear_page_index(dst_vma, dst_addr)= ; > > + } > > + > > orig_src_pte =3D ptep_get_and_clear(mm, src_addr, src_pte); > > set_pte_at(mm, dst_addr, dst_pte, orig_src_pte); > > - double_pt_unlock(dst_ptl, src_ptl); > > > > - return 0; > > +out: > > + double_pt_unlock(dst_ptl, src_ptl); > > + return err; > > } > > > > static int move_zeropage_pte(struct mm_struct *mm, > > @@ -1137,6 +1149,7 @@ static int move_pages_pte(struct mm_struct *mm, p= md_t *dst_pmd, pmd_t *src_pmd, > > __u64 mode) > > { > > swp_entry_t entry; > > + struct swap_info_struct *si =3D NULL; > > pte_t orig_src_pte, orig_dst_pte; > > pte_t src_folio_pte; > > spinlock_t *src_ptl, *dst_ptl; > > @@ -1318,6 +1331,8 @@ static int move_pages_pte(struct mm_struct *mm, p= md_t *dst_pmd, pmd_t *src_pmd, > > orig_dst_pte, orig_src_pte, dst_pm= d, > > dst_pmdval, dst_ptl, src_ptl, src_= folio); > > } else { > > + struct folio *folio =3D NULL; > > + > > entry =3D pte_to_swp_entry(orig_src_pte); > > if (non_swap_entry(entry)) { > > if (is_migration_entry(entry)) { > > @@ -1331,9 +1346,47 @@ static int move_pages_pte(struct mm_struct *mm, = pmd_t *dst_pmd, pmd_t *src_pmd, > > goto out; > > } > > > > - err =3D move_swap_pte(mm, dst_addr, src_addr, dst_pte, sr= c_pte, > > - orig_dst_pte, orig_src_pte, dst_pmd, > > - dst_pmdval, dst_ptl, src_ptl); > > + if (!pte_swp_exclusive(orig_src_pte)) { > > + err =3D -EBUSY; > > + goto out; > > + } > > + > > + si =3D get_swap_device(entry); > > + if (unlikely(!si)) { > > + err =3D -EAGAIN; > > + goto out; > > + } > > + /* > > + * Check if swapcache exists. If it does, the folio mu= st be > > + * moved even if the PTE is a swap entry. For large fo= lios, > > + * we directly return -EBUSY, as split_folio() current= ly > > + * also returns -EBUSY when attempting to split unmapp= ed > > + * large folios in the swapcache. This needs to be fix= ed > > + * to allow proper handling. > > + */ > > Some alignment issue on comments... > > We could also add something on the decision on why not taking anon_vma. > IIUC mention that no possible rmap walker when exclusive and unmapped > should be ok as of now.. > > > + if (!src_folio) > > + folio =3D filemap_get_folio(swap_address_space(en= try), > > + swap_cache_index(entry)); > > + if (!IS_ERR_OR_NULL(folio)) { > > + if (folio_test_large(folio)) { > > + err =3D -EBUSY; > > + folio_put(folio); > > + goto out; > > + } > > + src_folio =3D folio; > > We need to update src_folio_pte here, or later it might access > uninitialized stack var when a retry needed: > > if (src_folio && unlikely(!pte_same(src_folio_pte, orig_src_pte))= ) { > err =3D -EAGAIN; > goto out; > } > > > + if (!folio_trylock(src_folio)) { > > + pte_unmap(&orig_src_pte); > > + pte_unmap(&orig_dst_pte); > > + src_pte =3D dst_pte =3D NULL; > > + /* now we can block and wait */ > > + folio_lock(src_folio); > > + si =3D NULL; > > Swap device ref leak? > > Thanks, > > > + goto retry; > > + } > > + } > > + err =3D move_swap_pte(mm, dst_vma, dst_addr, src_addr, ds= t_pte, src_pte, > > + orig_dst_pte, orig_src_pte, dst_pmd, dst_= pmdval, > > + dst_ptl, src_ptl, src_folio); > > } > > > > out: > > @@ -1350,6 +1403,8 @@ static int move_pages_pte(struct mm_struct *mm, p= md_t *dst_pmd, pmd_t *src_pmd, > > if (src_pte) > > pte_unmap(src_pte); > > mmu_notifier_invalidate_range_end(&range); > > + if (si) > > + put_swap_device(si); > > > > return err; > > } > > > > If there are no objections, I'll send v2 tomorrow with the above code. > > 12:04 AM, Time to get some sleep now! :-) > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > My point is that if we want a proper fix for mTHP, we'd bette= r handle writeback. > > > > > > > Otherwise, this isn=E2=80=99t much different from directly re= turning -EBUSY as proposed > > > > > > > in this RFC. > > > > > > > > > > > > > > For small folios, there=E2=80=99s no split_folio issue, makin= g it relatively > > > > > > > simpler. Lokesh > > > > > > > mentioned plans to madvise NOHUGEPAGE in ART, so fixing small= folios is likely > > > > > > > the first priority. > > > > > > > > > > > > Agreed. > > > > > > > > > > > > -- > > > > > > Peter Xu > > > > > > > > > > Thanks > > Barry > > > > > > -- > Peter Xu >