From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from kanga.kvack.org (kanga.kvack.org [205.233.56.17]) by smtp.lore.kernel.org (Postfix) with ESMTP id 56C65C021B2 for ; Thu, 20 Feb 2025 22:34:11 +0000 (UTC) Received: by kanga.kvack.org (Postfix) id DF218280009; Thu, 20 Feb 2025 17:34:10 -0500 (EST) Received: by kanga.kvack.org (Postfix, from userid 40) id DA1D7280008; Thu, 20 Feb 2025 17:34:10 -0500 (EST) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id C1B3B280009; Thu, 20 Feb 2025 17:34:10 -0500 (EST) X-Delivered-To: linux-mm@kvack.org Received: from relay.hostedemail.com (smtprelay0010.hostedemail.com [216.40.44.10]) by kanga.kvack.org (Postfix) with ESMTP id 9CC90280008 for ; Thu, 20 Feb 2025 17:34:10 -0500 (EST) Received: from smtpin06.hostedemail.com (a10.router.float.18 [10.200.18.1]) by unirelay06.hostedemail.com (Postfix) with ESMTP id 1FDC7B6D93 for ; Thu, 20 Feb 2025 22:34:10 +0000 (UTC) X-FDA: 83141777460.06.3A29B8E Received: from mail-qt1-f182.google.com (mail-qt1-f182.google.com [209.85.160.182]) by imf01.hostedemail.com (Postfix) with ESMTP id 26BDE40010 for ; Thu, 20 Feb 2025 22:34:07 +0000 (UTC) Authentication-Results: imf01.hostedemail.com; dkim=pass header.d=google.com header.s=20230601 header.b=2NVCEv2V; spf=pass (imf01.hostedemail.com: domain of lokeshgidra@google.com designates 209.85.160.182 as permitted sender) smtp.mailfrom=lokeshgidra@google.com; dmarc=pass (policy=reject) header.from=google.com ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=hostedemail.com; s=arc-20220608; t=1740090848; h=from:from:sender:reply-to:subject:subject:date:date: message-id:message-id:to:to:cc:cc:mime-version:mime-version: content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references:dkim-signature; bh=s18TDU7nLAfv+jAsAfwI6kFUd8CxlMfAA6zR3ESNg4M=; b=WM54Vn6Ou0lXujbhe5yJuWCyPS2rGElu1yHfdvA5JISVEPlQv/6W4kh2jVcN0N3AudB2qi k0gQJrNO1tdJ2TG7TZ+gcSuYDE3udavshBK3KcIu649jewLFQIm+CvOqTk82H9FnCRVOyS O8B9ezmn0Gionmw2GJBcGY5A1Y/SufQ= ARC-Authentication-Results: i=1; imf01.hostedemail.com; dkim=pass header.d=google.com header.s=20230601 header.b=2NVCEv2V; spf=pass (imf01.hostedemail.com: domain of lokeshgidra@google.com designates 209.85.160.182 as permitted sender) smtp.mailfrom=lokeshgidra@google.com; dmarc=pass (policy=reject) header.from=google.com ARC-Seal: i=1; s=arc-20220608; d=hostedemail.com; t=1740090848; a=rsa-sha256; cv=none; b=bBZ069jmxSVesb0YAqrasUxPNhDT8lsUI2gW0bFecBTWwgd9K4cJtBzf4hcfVcNoWS+xiI 2q39jfk1hIkABu3MXLcVIHQxL8Ios/COZupi7JmWnwmjUbt2XAkFG57h98kRahcXPeVEmL hARWBXdu5PqblXtf8rbI5XP5yK3Bkjs= Received: by mail-qt1-f182.google.com with SMTP id d75a77b69052e-471fa3b19bcso87151cf.0 for ; Thu, 20 Feb 2025 14:34:07 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20230601; t=1740090847; x=1740695647; darn=kvack.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=s18TDU7nLAfv+jAsAfwI6kFUd8CxlMfAA6zR3ESNg4M=; b=2NVCEv2VuZEfWXAbmR2ytQfjPmKKlfUFcXYBRI4P2Tssm7Pi2HHyb03MkqZ5lVAHL2 eVnOTkmSgp3i+yHTyPNtnisvMVJYI/qRgnfF6SLI89oihBXBfwxKMMOOoT0QL0cpJdzb 0eumr9Rs1ssLB3+gO7K4Pv5IFxQpLOAqm+6ZVXmXuOp2q3hSgBwuBRZ583zVrkF0Rei4 a366+zgzNV0ttby73iEPrkxZ/QXVaTTFehBkDDLIWFHDMZ4ypQTFDzVAOVrhucPMsYx3 ac7CO1mOWYeea65WIekU2lQErA4gJWgp3oi9TIKmaYpIUJWXd5rwIGVS06eFM5ofOnXw 0P5g== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1740090847; x=1740695647; 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=s18TDU7nLAfv+jAsAfwI6kFUd8CxlMfAA6zR3ESNg4M=; b=CCFfqbwnB3UsiveR0Iujb4FuDNDppY1QGimb+pA0Wo48BnQMBg6ZMcRcORongTRW55 fc/jYqw55xqPFdB7dUuoA2uvj7Riqh9y8l/duLlO+Vn162MHtMPNHtnp37vWSKLzlgAS 9CIMLXfdC+0pr/vAfdlGdyVYLUhv+k+DNDseFOOiWPET87EY5j1tLBcm+p4jsEB+BkRo RXGo2ppo9WbMC8LfrQy/PbYlRbeJJmFbYMYm2B/nGqJlpnvjmv2nNiB/Sy0l/yE2t0ot CTBS9Sd5iLt3bzSb1tcHGA3NnQQdqQTvQ8a04b9GbJzFiDABq0Y7JfJ7Y/ClhV3L8nwK soug== X-Forwarded-Encrypted: i=1; AJvYcCWgvC10ffMLSn3E+sGb5MQ99VPI0xreAlbnaspBq4FRksuluS3xj8QLFAWie8ecu4aOY5goSPRvJA==@kvack.org X-Gm-Message-State: AOJu0YxDtia77tar0n8QFE0OM1YE1Uq4u4H1miebUp/n1KKCX385ENX3 q7BOGt7aVchthhMgbBB/9A43TNQdTEsK8ckAA6iXQHpE+8a/+/1QiwkMT86ZUP75zXpiUn9Qth+ JA6JODkKgrEK/BLzRgmvDiPZXPh2PbD0IdC0Y X-Gm-Gg: ASbGncsd6GgwotgzOMUSgpSiCiOmOIDuwR/3jCrzWKHJJprHZjAyFQqpUuwC6jaKfiA BASXJnmQOBbEeU5D7l1JgrxscydZjyxUcQu03CSH98ttwWn2VFJt8uYDIaEJ50O/+VhSD9SI+so b7JTSCMsFmINLwH8UWZjBNJ+nt8JcnVg== X-Google-Smtp-Source: AGHT+IFaKfQaSxINmYWnKKwUQzBgAQJiTmGrBos4qB/CMRfs2osPM0yAqO1xAg04RkqtaFkf0ZBL3BKRmBWXRH9yA0s= X-Received: by 2002:a05:622a:6084:b0:472:636:f614 with SMTP id d75a77b69052e-472238f26damr1239221cf.20.1740090846652; Thu, 20 Feb 2025 14:34:06 -0800 (PST) MIME-Version: 1.0 References: <20250219112519.92853-1-21cnbao@gmail.com> <50566d42-7754-4017-b290-f29d92e69231@redhat.com> In-Reply-To: From: Lokesh Gidra Date: Thu, 20 Feb 2025 14:33:55 -0800 X-Gm-Features: AWEUYZk-_BgvBJIYDchal1GgyBFXkQ69N3dx5i2RCijzL_0xnqI4iNpaSM4iTpQ Message-ID: Subject: Re: [PATCH RFC] mm: Fix kernel BUG when userfaultfd_move encounters swapcache To: Barry Song <21cnbao@gmail.com> Cc: David Hildenbrand , Suren Baghdasaryan , linux-mm@kvack.org, akpm@linux-foundation.org, linux-kernel@vger.kernel.org, zhengtangquan@oppo.com, Barry Song , Andrea Arcangeli , Al Viro , Axel Rasmussen , Brian Geffon , Christian Brauner , Hugh Dickins , Jann Horn , Kalesh Singh , "Liam R . Howlett" , Matthew Wilcox , Michal Hocko , Mike Rapoport , Nicolas Geoffray , Peter Xu , Ryan Roberts , Shuah Khan , ZhangPeng , Yu Zhao Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable X-Rspam-User: X-Rspamd-Server: rspam08 X-Rspamd-Queue-Id: 26BDE40010 X-Stat-Signature: nn17up4ithwwycu1u3cy84uxpmgro5s4 X-HE-Tag: 1740090847-166926 X-HE-Meta: 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 EWg9GLEJ 5bgnNM69EFO3EWLN+X5e28VDd/lsg5zqIh0TnWqv9Lb9zjJZeyZPFD8zsFmF971rbOqNjGTyAcwpLsBv3ooRkNgjkfvxQB1nUtsXkliM0gIx5+vC9Zb8Mcyv9BWdiuRH/KbBXLo0JzbC1152Cj3cKdpjucVFo8ErHmvY6K8llL30LyqDy+CuKcmMZDk2+Y7gcmomUALYSVMBab5ELZfBjX+Yi5MKg34ivDy8j0WC4xy24eeF7AL8ClksLMWg+cdFNRkAFZf8FfzXy4Bq7xkS9nPQMV1E5bgEmvO2/qY6hRg0HJRpZNIX/IC3Am50dcD5RichyLzlJBJ/X3k2ru3sJkF3M1KMLyIpRK0cL0jYTT8uqNn8= X-Bogosity: Ham, tests=bogofilter, spamicity=0.000000, version=1.2.4 Sender: owner-linux-mm@kvack.org Precedence: bulk X-Loop: owner-majordomo@kvack.org List-ID: List-Subscribe: List-Unsubscribe: On Thu, Feb 20, 2025 at 2:27=E2=80=AFPM Barry Song <21cnbao@gmail.com> wrot= e: > > On Fri, Feb 21, 2025 at 11:20=E2=80=AFAM Lokesh Gidra wrote: > > > > On Thu, Feb 20, 2025 at 1:45=E2=80=AFPM Barry Song <21cnbao@gmail.com> = wrote: > > > > > > On Thu, Feb 20, 2025 at 10:36=E2=80=AFPM David Hildenbrand wrote: > > > > > > > > On 20.02.25 10:31, Barry Song wrote: > > > > > On Thu, Feb 20, 2025 at 9:51=E2=80=AFPM David Hildenbrand wrote: > > > > >> > > > > >> On 19.02.25 21:37, Barry Song wrote: > > > > >>> On Thu, Feb 20, 2025 at 7:27=E2=80=AFAM Suren Baghdasaryan wrote: > > > > >>>> > > > > >>>> On Wed, Feb 19, 2025 at 3:25=E2=80=AFAM Barry Song <21cnbao@gm= ail.com> wrote: > > > > >>>>> > > > > >>>>> From: Barry Song > > > > >>>>> > > > > >>>>> userfaultfd_move() checks whether the PTE entry is present or= a > > > > >>>>> swap entry. > > > > >>>>> > > > > >>>>> - If the PTE entry is present, move_present_pte() handles fol= io > > > > >>>>> 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() simply co= pies > > > > >>>>> 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 swa= p > > > > >>>>> cache. > > > > >>>>> > > > > >>>>> If do_swap_page() is triggered, it may locate the folio in th= e > > > > >>>>> swap cache. However, during add_rmap operations, a kernel pan= ic > > > > >>>>> can occur due to: > > > > >>>>> page_pgoff(folio, page) !=3D linear_page_index(vma, addres= s) > > > > >>>> > > > > >>>> Thanks for the report and reproducer! > > > > >>>> > > > > >>>>> > > > > >>>>> $./a.out > /dev/null > > > > >>>>> [ 13.336953] page: refcount:6 mapcount:1 mapping:00000000f4= 3db19c index:0xffffaf150 pfn:0x4667c > > > > >>>>> [ 13.337520] head: order:2 mapcount:1 entire_mapcount:0 nr_= pages_mapped:1 pincount:0 > > > > >>>>> [ 13.337716] memcg:ffff00000405f000 > > > > >>>>> [ 13.337849] anon flags: 0x3fffc0000020459(locked|uptodate|= dirty|owner_priv_1|head|swapbacked|node=3D0|zone=3D0|lastcpupid=3D0xffff) > > > > >>>>> [ 13.338630] raw: 03fffc0000020459 ffff80008507b538 ffff800= 08507b538 ffff000006260361 > > > > >>>>> [ 13.338831] raw: 0000000ffffaf150 0000000000004000 0000000= 600000000 ffff00000405f000 > > > > >>>>> [ 13.339031] head: 03fffc0000020459 ffff80008507b538 ffff80= 008507b538 ffff000006260361 > > > > >>>>> [ 13.339204] head: 0000000ffffaf150 0000000000004000 000000= 0600000000 ffff00000405f000 > > > > >>>>> [ 13.339375] head: 03fffc0000000202 fffffdffc0199f01 ffffff= ff00000000 0000000000000001 > > > > >>>>> [ 13.339546] head: 0000000000000004 0000000000000000 000000= 00ffffffff 0000000000000000 > > > > >>>>> [ 13.339736] page dumped because: VM_BUG_ON_PAGE(page_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: 00000000f2000800 [= #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 -DI= T -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: fffffdffc0199f00 x2= 7: 0000000000000001 > > > > >>>>> [ 13.342404] x26: 0000000000000000 x25: 0000000000000001 x2= 4: 0000000000000001 > > > > >>>>> [ 13.342575] x23: 0000ffffaf0d0000 x22: 0000ffffaf0d0000 x2= 1: fffffdffc0199f00 > > > > >>>>> [ 13.342731] x20: fffffdffc0199f00 x19: ffff000006210700 x1= 8: 00000000ffffffff > > > > >>>>> [ 13.342881] x17: 6c203d2120296567 x16: 6170202c6f696c6f x1= 5: 662866666f67705f > > > > >>>>> [ 13.343033] x14: 6567617028454741 x13: 2929737365726464 x1= 2: ffff800083728ab0 > > > > >>>>> [ 13.343183] x11: ffff800082996bf8 x10: 0000000000000fd7 x9= : ffff80008011bc40 > > > > >>>>> [ 13.343351] x8 : 0000000000017fe8 x7 : 00000000fffff000 x6= : ffff8000829eebf8 > > > > >>>>> [ 13.343498] x5 : c0000000fffff000 x4 : 0000000000000000 x3= : 0000000000000000 > > > > >>>>> [ 13.343645] x2 : 0000000000000000 x1 : ffff0000062db980 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 97ff5779 (d42= 10000) > > > > >>>>> [ 13.345504] ---[ end trace 0000000000000000 ]--- > > > > >>>>> [ 13.345715] note: a.out[107] exited with irqs disabled > > > > >>>>> [ 13.345954] note: a.out[107] exited with preempt_count 2 > > > > >>>>> > > > > >>>>> Fully fixing it would be quite complex, requiring similar han= dling > > > > >>>>> 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 individually, > > > > >>> 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 for mov= e_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 after = 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(= )? > > > > >>> > > > > >>> By the way, I have also reported that userfaultfd_move() has a = fundamental > > > > >>> conflict with TAO (Cc'ed Yu Zhao), which has been part of the A= ndroid common > > > > >>> kernel. In this scenario, folios in the virtual zone won=E2=80= =99t be split in > > > > >>> split_folio(). Instead, the large folio migrates into nr_pages = small folios. > > > > >> > > Thus, the best-case scenario would be: > > > > >>> > > > > >>> mTHP -> migrate to small folios in split_folio() -> move small = folios to > > > > >>> dst_addr > > > > >>> > > > > >>> While this works, it negates the performance benefits of > > > > >>> userfaultfd_move(), as it introduces two PTE operations (migrat= ion in > > > > >>> split_folio() and move in userfaultfd_move() while retry), nr_p= ages memory > > > > >>> allocations, and still requires one memcpy(). This could end up > > > > >>> performing even worse than userfaultfd_copy(), I guess. > > > > >> > > The worst-case scenario would be failing to allocate small= folios in > > > > >>> split_folio(), then userfaultfd_move() might return -ENOMEM? > > > > >> > > > > >> Although that's an Android problem and not an upstream problem, = I'll > > > > >> note that there are other reasons why the split / move might fai= l, and > > > > >> user space either must retry or fallback to a COPY. > > > > >> > > > > >> Regarding mTHP, we could move the whole folio if the user space-= provided > > > > >> range allows for batching over multiple PTEs (nr_ptes), they are= in a > > > > >> single VMA, and folio_mapcount() =3D=3D nr_ptes. > > > > >> > > > > >> There are corner cases to handle, such as moving mTHPs such that= they > > > > >> suddenly cross two page tables I assume, that are harder to hand= le when > > > > >> not moving individual PTEs where that cannot happen. > > > > > > > > > > This is a useful suggestion. I=E2=80=99ve heard that Lokesh is al= so interested in > > > > > modifying ART to perform moves at the mTHP granularity, which wou= ld require > > > > > kernel modifications as well. It=E2=80=99s likely the direction w= e=E2=80=99ll take after > > > > > fixing the current urgent bugs. The current split_folio() really = isn=E2=80=99t ideal. > > > > > > > > > > The corner cases you mentioned are definitely worth considering. = However, > > > > > once we can perform batch UFFDIO_MOVE, I believe that in most cas= es, > > > > > the conflict between userfaultfd_move() and TAO will be resolved = ? > > > > > > > > Well, as soon as you would have varying mTHP sizes, you'd still run= into > > > > the split with TAO. Maybe that doesn't apply with Android today, bu= t I > > > > can just guess that performing sub-mTHP moving would still be requi= red > > > > for GC at some point. > > > > > > With patch v2[1], as discussed in my previous email, I have observed = that > > > small folios consistently succeed without crashing. Similarly, mTHP n= o > > > longer crashes; however, it still returns -EBUSY during the raced tim= e > > > window, even after adding folio_wait_writeback. While I previously > > > mentioned that folio_writeback prevents mTHP from splitting, this is = not > > > the only factor. The split_folio() function still returns -EBUSY beca= use > > > folio_get_anon_vma(folio) returns NULL when the folio is not mapped. > > > > > > int split_huge_page_to_list_to_order(struct page *page, struct list_h= ead *list, > > > unsigned int new_order) > > > { > > > anon_vma =3D folio_get_anon_vma(folio); > > > if (!anon_vma) { > > > ret =3D -EBUSY; > > > goto out; > > > } > > > > > > end =3D -1; > > > mapping =3D NULL; > > > anon_vma_lock_write(anon_vma); > > > } > > > > > > Even if mTHP is not from TAO's virtual zone, userfaultfd_move() will = still > > > fail when performing sub-mTHP moving in the swap cache case due to: > > > > Just to clarify my doubt. What do you mean by sub-mTHP? Also when you > > say 'small folio' above, do you mean single-page folios? > > This means any moving size smaller than the size of mTHP, or moving > a partial mTHP. > > > > > Am I understanding correctly that your patch correctly handles moving > > single swap-cache page case? > > Yes, the crash is fixed for both small and large folios, and for small > folios, moving is consistently successful(even for the swapcache case). > The only issue is that sub-mTHP moving constantly fails for the swapcache > case because split_folio() fails, even after waiting for writeback as > split_folio() > can only split mapped folios - which is false for swapcache since > try_to_unmap_one() has been done. > > So I'd say for mTHP, returning -EBUSY as early as possible is the > better choice to avoid wasting much time and eventually returning > -EBUSY anyway unless we want to modify split_folio() things. > Great! In this case, can we please fix the kernel panic bug as soon as possible. Until that is fixed, the ioctl is practically unusable. > > > > > > struct anon_vma *folio_get_anon_vma(const struct folio *folio) > > > { > > > ... > > > if (!folio_mapped(folio)) > > > goto out; > > > ... > > > } > > > > > > We likely need to modify split_folio() to support splitting unmapped = anon > > > folios within the swap cache or introduce a new function like > > > split_unmapped_anon_folio()? Otherwise, userspace will have to fall b= ack > > > to UFFDIO_COPY or retry. > > > > > > As it stands, I see no way for sub-mTHP to survive moving with the cu= rrent > > > code and within the existing raced window. For mTHP, there is essenti= ally > > > no difference between returning -EBUSY immediately upon detecting tha= t it > > > is within the swap cache, as proposed in v1. > > > > > > [1] https://lore.kernel.org/linux-mm/20250220092101.71966-1-21cnbao@g= mail.com/ > > > > > > > > > > > -- > > > > Cheers, > > > > > > > > David / dhildenb > > > > > > > > Thanks > Barry