From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mail-yx2-f42.google.com (mail-yx2-f42.google.com [74.125.224.170]) (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 E7B8D2F7EF4 for ; Mon, 5 Oct 2026 19:16:43 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=74.125.224.170 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1791227805; cv=none; b=PPJuMCAT/m5az6hs4jx+PlKCqpOZV4Fs7slSX+HzlVNiWF8C+b/Z5MfOKQ8/64zsqce1H/tle901pgPDsxarVSnneEpYM6vavB0mAhb4hSbY6Q3fYBqhqdo+LxdJ6D0Uuwbl38GPLb31ZOQIBZIa0qXUXOWvfAZK7w7FLFaVaZA= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1791227805; c=relaxed/simple; bh=ulQA5cvJgKQcqUv21AfRSI+EJFD+UJHNlzoqx4tqBmU=; h=Message-ID:Subject:From:To:Cc:Date:In-Reply-To:References: Content-Type:MIME-Version; b=nSizwGndYpu0BWbMAedPS3SgkCIMgzUoVSji5M7QPNkFwAHw2As9h7zjKZws5hvkAlGc8sn1RJDKljmrXcO6G3gsVfcULYmbXomMCwE1PJfaNVDGtzRa+EcfcPZrXwFy5Td6stabPlfyIKd7HzHYtEQr5EobsqME7uwNe/qeZs4= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dmarc=none (p=none dis=none) header.from=dubeyko.com; spf=pass smtp.mailfrom=dubeyko.com; dkim=pass (2048-bit key) header.d=dubeyko-com.20251104.gappssmtp.com header.i=@dubeyko-com.20251104.gappssmtp.com header.b=GwSVEqYW; arc=none smtp.client-ip=74.125.224.170 Authentication-Results: smtp.subspace.kernel.org; dmarc=none (p=none dis=none) header.from=dubeyko.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=dubeyko.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=dubeyko-com.20251104.gappssmtp.com header.i=@dubeyko-com.20251104.gappssmtp.com header.b="GwSVEqYW" Received: by mail-yx2-f42.google.com with SMTP id 956f58d0204a3-672f4b7091dso844540d50.0 for ; Mon, 05 Oct 2026 12:16:43 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=dubeyko-com.20251104.gappssmtp.com; s=20251104; t=1791227803; x=1791832603; darn=vger.kernel.org; h=mime-version:user-agent:content-transfer-encoding:content-type :autocrypt:references:in-reply-to:date:cc:to:from:subject:message-id :from:to:cc:subject:date:message-id:reply-to:content-type; bh=AD8CeJIlG5hCbH4+fGlxBDHvP5Fa9LzqNq1Ik5KKPnE=; b=GwSVEqYWtHHC96q7n4WUPWI9F3MOO5GW3DMhHtagv6MLw8qVPMjkauq5R+GGBZaY86 Xl4vVyFH5K92wgp/wPw91ZdwWKmr44VU+in5rQrg0MQxWTIVN9t20mdIkpaYSDIhnQWA QrMYQAS3MdyHUrLaW18G1G0Rl256DksndlZOnUwpwqRGsBDDnpzFxypdN2D9xmswXrGE wiI7H+WkcIhxnMQSWWarBslI4jK0EIT1W4ABcD1cTvq8f697JYwcA4fxq0CddaRgMHXL lVfqyUr0CeCsAfqq6GBzCN166EUFMxvpeAFMEDE44ecVfF7T2LAY2efctkjvM4UwDU40 tCxw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20260707; t=1791227803; x=1791832603; h=mime-version:user-agent:content-transfer-encoding:content-type :autocrypt:references:in-reply-to:date:cc:to:from:subject:message-id :x-gm-gg:x-gm-message-state:from:to:cc:subject:date:message-id :reply-to:content-type; bh=AD8CeJIlG5hCbH4+fGlxBDHvP5Fa9LzqNq1Ik5KKPnE=; b=syoLHuORiJWFlyqN1c3fk+EQXBRKU7CNbM2iXgL/yMsXzsi7agEi4JL7EjmySgHplt x6l1VCxRuAtb5BjZomFmJT1l3wsRpLfiA5o71JUCI7bM5IYg9j1U51LTa/eY03zcRQdm YYVNgkVl3vS2Nr4rIWCgWXwmce0119rqvPDmdaiYGj76YM0x5MCcGKjzK8TAJ3Nj20Dy VYRiUO2v8ewY45pGz/fIwTuwGcQBUDJYDzoidyQbdiPSM6gIzALBrIPqM+5UeDrh4FjQ eJpPIEKSFWjAoESyMyXIK5osrvuiechDCMVIXMfI8PDjPf2kZJa/tdA2ku0SuBqtk6Tu 2Ong== X-Forwarded-Encrypted: i=1; AKwUvBybOkBmkT72Sep5+/uChuD/e4o3o3HzsOubfKjX5DvdqdRqxIpHVoo2CAuM7PjOfQYO1URGfjm5/Qbk5ck=@vger.kernel.org X-Gm-Message-State: AFq9FYLZiVynp2PtCyVI7pkGI15BBksA0HGkHjCRhWRv9KumiJqEZz4N px1QTHeoxiaPtXOjuNVnzdBK2i4jUPKBIfqM19d3M+9Gp6iT6mRK/03VjzV0BO/Bs2M= X-Gm-Gg: AYBFou0Rdn01DVWYTwL5nIcXqkWidrrvd/mjnBO649GLvAV6FuglCvcoQFKIAj8TRCp URSUMOVgx4dxdLapmUW5iXJnwdAlAOWMxY1s/gxjUgPxr6OWz8yrhXRY5nnMTf4smImsDiEE38b o83KiDb1C7HuGpmoovzkS6S2s/Td8Axi31/FKzzw0XvmsqOoikD+aK3lTfhQqy2pLKGsDeIQslh 9TGA+EiI60dbhjM9x52dXvStVO03hic+G4QXd8vfj8VFt+/K601NoguFLM2vSPxIUwS7m5XwCUl 61wlDz7oHmyt/VLoCRGvHzQ5v96EbvC4wPTeUJ3gvazdUQNp1kd6U9WGt9gFqUJN8CgXTe+Fy9l AkHl1DORprQVOqRKcit/gLs8OjJaeHQvIodS7Ina8fivyEH38e5XRfNsmxslrE0Nql4MBKKMzk7 griE9N1pu6wXWXLmIZRsLd7LFkhZWkHxUp71TEADd4y0bHOqNpNBK2bmWfHn6uFXWpbcv+xqw8E XWLChfetNPXRw0emwamNxQlxTifsfvbdj/h//5ZtgpajRp0tLA4YKKS+bf3iE/G97X6HqW2AuwB hSD0o9XziCH5kp8Y/GrnrLWnxRLrdneAbAhY54FOsRjeM2jiT1WBtjC4FJw1nsT0 X-Received: by 2002:a05:690e:4093:b0:675:449d:3d72 with SMTP id 956f58d0204a3-677f15a6050mr281171d50.32.1791227802743; Mon, 05 Oct 2026 12:16:42 -0700 (PDT) Received: from ?IPv6:2600:1700:6476:1430:3aa0:911f:4e09:b45d? ([2600:1700:6476:1430:3aa0:911f:4e09:b45d]) by smtp.gmail.com with ESMTPSA id 956f58d0204a3-677c1b1a214sm3304995d50.6.2026.10.05.12.16.38 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Mon, 05 Oct 2026 12:16:40 -0700 (PDT) Message-ID: <0eb39e561e4c81c0fe2198a2ba352ba81f8d94b7.camel@dubeyko.com> Subject: Re: [PATCH v3 1/2] hfsplus: free cached B-tree nodes on hfs_btree_open() error path From: Viacheslav Dubeyko To: Mahmut Emin Kurhan , linux-fsdevel@vger.kernel.org Cc: glaubitz@physik.fu-berlin.de, frank.li@vivo.com, linux-kernel@vger.kernel.org Date: Mon, 05 Oct 2026 12:16:37 -0700 In-Reply-To: <20261001213807.1519745-2-guvenlik@noroxi.com> References: <0de056179e3e0a7c4ef525cdb143647b7493dc71.camel@dubeyko.com> <20261001213807.1519745-1-guvenlik@noroxi.com> <20261001213807.1519745-2-guvenlik@noroxi.com> Autocrypt: addr=slava@dubeyko.com; prefer-encrypt=mutual; keydata=mQINBGgaTLYBEADaJc/WqWTeunGetXyyGJ5Za7b23M/ozuDCWCp+yWUa2GqQKH40dxRIR zshgOmAue7t9RQJU9lxZ4ZHWbi1Hzz85+0omefEdAKFmxTO6+CYV0g/sapU0wPJws3sC2Pbda9/eJ ZcvScAX2n/PlhpTnzJKf3JkHh3nM1ACO3jzSe2/muSQJvqMLG2D71ccekr1RyUh8V+OZdrPtfkDam V6GOT6IvyE+d+55fzmo20nJKecvbyvdikWwZvjjCENsG9qOf3TcCJ9DDYwjyYe1To8b+mQM9nHcxp jUsUuH074BhISFwt99/htZdSgp4csiGeXr8f9BEotRB6+kjMBHaiJ6B7BIlDmlffyR4f3oR/5hxgy dvIxMocqyc03xVyM6tA4ZrshKkwDgZIFEKkx37ec22ZJczNwGywKQW2TGXUTZVbdooiG4tXbRBLxe ga/NTZ52ZdEkSxAUGw/l0y0InTtdDIWvfUT+WXtQcEPRBE6HHhoeFehLzWL/o7w5Hog+0hXhNjqte fzKpI2fWmYzoIb6ueNmE/8sP9fWXo6Av9m8B5hRvF/hVWfEysr/2LSqN+xjt9NEbg8WNRMLy/Y0MS p5fgf9pmGF78waFiBvgZIQNuQnHrM+0BmYOhR0JKoHjt7r5wLyNiKFc8b7xXndyCDYfniO3ljbr0j tXWRGxx4to6FwARAQABtCZWaWFjaGVzbGF2IER1YmV5a28gPHNsYXZhQGR1YmV5a28uY29tPokCVw QTAQoAQQIbAQUJA8JnAAULCQgHAgYVCgkICwIEFgIDAQIeAQIXgBYhBFXDC2tnzsoLQtrbBDlc2cL fhEB1BQJoGl5PAhkBAAoJEDlc2cLfhEB17DsP/jy/Dx19MtxWOniPqpQf2s65enkDZuMIQ94jSg7B F2qTKIbNR9SmsczjyjC+/J7m7WZRmcqnwFYMOyNfh12aF2WhjT7p5xEAbvfGVYwUpUrg/lcacdT0D Yk61GGc5ZB89OAWHLr0FJjI54bd7kn7E/JRQF4dqNsxU8qcPXQ0wLHxTHUPZu/w5Zu/cO+lQ3H0Pj pSEGaTAh+tBYGSvQ4YPYBcV8+qjTxzeNwkw4ARza8EjTwWKP2jWAfA/ay4VobRfqNQ2zLoo84qDtN Uxe0zPE2wobIXELWkbuW/6hoQFPpMlJWz+mbvVms57NAA1HO8F5c1SLFaJ6dN0AQbxrHi45/cQXla 9hSEOJjxcEnJG/ZmcomYHFneM9K1p1K6HcGajiY2BFWkVet9vuHygkLWXVYZ0lr1paLFR52S7T+cf 6dkxOqu1ZiRegvFoyzBUzlLh/elgp3tWUfG2VmJD3lGpB3m5ZhwQ3rFpK8A7cKzgKjwPp61Me0o9z HX53THoG+QG+o0nnIKK7M8+coToTSyznYoq9C3eKeM/J97x9+h9tbizaeUQvWzQOgG8myUJ5u5Dr4 6tv9KXrOJy0iy/dcyreMYV5lwODaFfOeA4Lbnn5vRn9OjuMg1PFhCi3yMI4lA4umXFw0V2/OI5rgW BQELhfvW6mxkihkl6KLZX8m1zcHitCpWaWFjaGVzbGF2IER1YmV5a28gPFNsYXZhLkR1YmV5a29Aa WJtLmNvbT6JAlQEEwEKAD4WIQRVwwtrZ87KC0La2wQ5XNnC34RAdQUCaBpd7AIbAQUJA8JnAAULCQ gHAgYVCgkICwIEFgIDAQIeAQIXgAAKCRA5XNnC34RAdYjFEACiWBEybMt1xjRbEgaZ3UP5i2bSway DwYDvgWW5EbRP7JcqOcZ2vkJwrK3gsqC3FKpjOPh7ecE0I4vrabH1Qobe2N8B2Y396z24mGnkTBbb 16Uz3PC93nFN1BA0wuOjlr1/oOTy5gBY563vybhnXPfSEUcXRd28jI7z8tRyzXh2tL8ZLdv1u4vQ8 E0O7lVJ55p9yGxbwgb5vXU4T2irqRKLxRvU80rZIXoEM7zLf5r7RaRxgwjTKdu6rYMUOfoyEQQZTD 4Xg9YE/X8pZzcbYFs4IlscyK6cXU0pjwr2ssjearOLLDJ7ygvfOiOuCZL+6zHRunLwq2JH/RmwuLV mWWSbgosZD6c5+wu6DxV15y7zZaR3NFPOR5ErpCFUorKzBO1nA4dwOAbNym9OGkhRgLAyxwpea0V0 ZlStfp0kfVaSZYo7PXd8Bbtyjali0niBjPpEVZdgtVUpBlPr97jBYZ+L5GF3hd6WJFbEYgj+5Af7C UjbX9DHweGQ/tdXWRnJHRzorxzjOS3003ddRnPtQDDN3Z/XzdAZwQAs0RqqXrTeeJrLppFUbAP+HZ TyOLVJcAAlVQROoq8PbM3ZKIaOygjj6Yw0emJi1D9OsN2UKjoe4W185vamFWX4Ba41jmCPrYJWAWH fAMjjkInIPg7RLGs8FiwxfcpkILP0YbVWHiNAabQoVmlhY2hlc2xhdiBEdWJleWtvIDx2ZHViZXlr b0BrZXJuZWwub3JnPokCVAQTAQoAPhYhBFXDC2tnzsoLQtrbBDlc2cLfhEB1BQJoVemuAhsBBQkDw mcABQsJCAcCBhUKCQgLAgQWAgMBAh4BAheAAAoJEDlc2cLfhEB1GRwP/1scX5HO9Sk7dRicLD/fxo ipwEs+UbeA0/TM8OQfdRI4C/tFBYbQCR7lD05dfq8VsYLEyrgeLqP/iRhabLky8LTaEdwoAqPDc/O 9HRffx/faJZqkKc1dZryjqS6b8NExhKOVWmDqN357+Cl/H4hT9wnvjCj1YEqXIxSd/2Pc8+yw/KRC AP7jtRzXHcc/49Lpz/NU5irScusxy2GLKa5o/13jFK3F1fWX1wsOJF8NlTx3rLtBy4GWHITwkBmu8 zI4qcJGp7eudI0l4xmIKKQWanEhVdzBm5UnfyLIa7gQ2T48UbxJlWnMhLxMPrxgtC4Kos1G3zovEy Ep+fJN7D1pwN9aR36jVKvRsX7V4leIDWGzCdfw1FGWkMUfrRwgIl6i3wgqcCP6r9YSWVQYXdmwdMu 1RFLC44iF9340S0hw9+30yGP8TWwd1mm8V/+zsdDAFAoAwisi5QLLkQnEsJSgLzJ9daAsE8KjMthv hUWHdpiUSjyCpigT+KPl9YunZhyrC1jZXERCDPCQVYgaPt+Xbhdjcem/ykv8UVIDAGVXjuk4OW8la nf8SP+uxkTTDKcPHOa5rYRaeNj7T/NClRSd4z6aV3F6pKEJnEGvv/DFMXtSHlbylhyiGKN2Amd0b4 9jg+DW85oNN7q2UYzYuPwkHsFFq5iyF1QggiwYYTpoVXsw Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable User-Agent: Evolution 3.60.1 (by Flathub.org) Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 On Thu, 2026-10-01 at 23:38 +0200, Mahmut Emin Kurhan wrote: > hfs_btree_open() can fail after hfs_bnode_find(tree, > HFSPLUS_TREE_HEAD) has > already inserted the head node into tree->node_hash. >=20 > __hfs_bnode_create() inserts the new bnode into tree->node_hash > before it > reads the node pages; if a page read fails it sets HFS_BNODE_ERROR > and > returns the node still hashed. hfs_bnode_find() then takes its > node_error > path and calls hfs_bnode_put(), which only frees a node once its > refcount > reaches zero *and* HFS_BNODE_DELETED is set. For the errored head > node that > flag is not set, so it stays in tree->node_hash with a zero refcount. >=20 > hfs_btree_open() then sees IS_ERR(node) and jumps to free_tree:, > doing a > bare kfree(tree). Only hfs_btree_close() walks tree->node_hash[] and > frees > the cached nodes, so the head node is leaked. Mounting a crafted HFS+ > image > whose head B-tree node fails to read leaks kernel memory on every > attempt. >=20 > Factor the node-hash freeing out of hfs_btree_close() into a small > helper > hfs_bnode_hash_free() and call it from both hfs_btree_close() and the > hfs_btree_open() error path. The helper keeps the hash_lock > serialization > added in commit dce0e0248205 ("hfs/hfsplus: serialize B-tree close I think it is not good idea of mentioning this. This patch is not upstream yet. > against > folio release"); on the open error path the tree has not been > published yet, > so the lock is uncontended but harmless. >=20 > Found via coverage-guided fuzzing (syzkaller + kmemleak) by Noroxi. Should we use Reported-by: tag? >=20 > Signed-off-by: Mahmut Emin Kurhan > --- > =C2=A0fs/hfsplus/btree.c | 41 +++++++++++++++++++++++------------------ > =C2=A01 file changed, 23 insertions(+), 18 deletions(-) >=20 > diff --git a/fs/hfsplus/btree.c b/fs/hfsplus/btree.c > index bd4dbdbc8f..2183828d5d 100644 > --- a/fs/hfsplus/btree.c > +++ b/fs/hfsplus/btree.c > @@ -265,6 +265,27 @@ static const char *hfs_btree_name(u32 cnid) > =C2=A0} > =C2=A0 > =C2=A0/* Get a reference to a B*Tree and do some initial checks */ Are you sure that it is correct place for the comment? > +static void hfs_bnode_hash_free(struct hfs_btree *tree) Maybe, hfs_btree_free_nodes() because function lives in btree.c? > +{ > + struct hfs_bnode *node; > + int i; > + > + for (i =3D 0; i < NODE_HASH_SIZE; i++) { > + spin_lock(&tree->hash_lock); > + while ((node =3D tree->node_hash[i])) { > + hfs_bnode_unhash(node); > + spin_unlock(&tree->hash_lock); > + if (atomic_read(&node->refcnt)) > + pr_crit("node %d:%d still has %d > user(s)!\n", > + node->tree->cnid, node- > >this, > + atomic_read(&node->refcnt)); > + hfs_bnode_free(node); > + spin_lock(&tree->hash_lock); > + } > + spin_unlock(&tree->hash_lock); > + } > +} > + > =C2=A0struct hfs_btree *hfs_btree_open(struct super_block *sb, u32 id) > =C2=A0{ > =C2=A0 struct hfs_btree *tree; > @@ -403,6 +424,7 @@ struct hfs_btree *hfs_btree_open(struct > super_block *sb, u32 id) > =C2=A0 tree->inode->i_mapping->a_ops =3D &hfsplus_aops; > =C2=A0 iput(tree->inode); > =C2=A0 free_tree: > + hfs_bnode_hash_free(tree); We have inconsistency here. The hfs_btree_close() calls hfs_bnode_hash_free() and iput() then. Here we have opposite sequence. Thanks, Slava. > =C2=A0 kfree(tree); > =C2=A0 return NULL; > =C2=A0} > @@ -410,27 +432,10 @@ struct hfs_btree *hfs_btree_open(struct > super_block *sb, u32 id) > =C2=A0/* Release resources used by a btree */ > =C2=A0void hfs_btree_close(struct hfs_btree *tree) > =C2=A0{ > - struct hfs_bnode *node; > - int i; > - > =C2=A0 if (!tree) > =C2=A0 return; > =C2=A0 > - for (i =3D 0; i < NODE_HASH_SIZE; i++) { > - spin_lock(&tree->hash_lock); > - while ((node =3D tree->node_hash[i])) { > - hfs_bnode_unhash(node); > - spin_unlock(&tree->hash_lock); > - if (atomic_read(&node->refcnt)) > - pr_crit("node %d:%d " > - "still has %d > user(s)!\n", > - node->tree->cnid, node- > >this, > - atomic_read(&node->refcnt)); > - hfs_bnode_free(node); > - spin_lock(&tree->hash_lock); > - } > - spin_unlock(&tree->hash_lock); > - } > + hfs_bnode_hash_free(tree); > =C2=A0 iput(tree->inode); > =C2=A0 kfree(tree); > =C2=A0}