From nobody Sun Oct 5 12:52:24 2025 Received: from mail-pj1-f74.google.com (mail-pj1-f74.google.com [209.85.216.74]) (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 BA610291864 for ; Mon, 4 Aug 2025 23:15:58 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.216.74 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1754349360; cv=none; b=K/RJ5IDvzM1A2mcSDZAYvvMBbD6kc7ytypaRJApiZr+aamg5Kui6IonbBPLHmLyIhSXvOlpVTFWS+rOABNon98FjViDyVPWYpYIZBkOUpyQ/yGV8DaQVonCX+T+HC3pEqlUUP7B8VestiwKZpzLhbUTzNxRNUS2dfQ85v0VhSo0= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1754349360; c=relaxed/simple; bh=JLzkhn9Ai/ASHzkFTaBGXCUct2ehP9RlVk17yOUGn80=; h=Date:In-Reply-To:Mime-Version:References:Message-ID:Subject:From: To:Cc:Content-Type; b=NdH8fhQ2jT0ChIHelQ5xpu7iwyxpxIQS9I8lE53jJZu8dYcjdnmGSZ1skb8pNno6L4koGg+7HLf3jUUrXtFYQV+oiezbW7olUKI3BtIwBvB4lB+42duadtFC3uxLhkXDhmL8bNxjLGYMFJA6TB2H5u4hmf9hkLOGMrpNjyKJm0Q= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com; spf=pass smtp.mailfrom=flex--surenb.bounces.google.com; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b=DQ30KFxq; arc=none smtp.client-ip=209.85.216.74 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=flex--surenb.bounces.google.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b="DQ30KFxq" Received: by mail-pj1-f74.google.com with SMTP id 98e67ed59e1d1-31f74a64da9so5552359a91.2 for ; Mon, 04 Aug 2025 16:15:58 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20230601; t=1754349358; x=1754954158; darn=vger.kernel.org; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:from:to:cc:subject:date:message-id:reply-to; bh=rATS4q8eQgXRIMxeVI6i3f0N8VShHmldfKu+DIDZEpI=; b=DQ30KFxqYOaZXRNTrSxlXldpQUcqbHwHSgynKNa3HXIMdBYFt/K/ixVcgA6JEYMJg7 6Gm5SXGZdxv2vHFnETBlP+qy5KPhCz6LX/TlAcweeesiDf6grkyvWSwk+f6UrBwxGeDB SZojkVFr50Uv477pjsSSKQyUmfFGg1eSFSUM4kqrqfpZYEGPE7PS9aUoJqCFQbnyrDI9 i0985pUjNZXfyDCoCFOw+zAM4VVbzVMMmh1trDJ7uOTARtGxblcrJfBZTbvGg3Dds8LZ d/29d1zT+mlXG/XIy85kLjkOyqiJg+OOQf12aKGxBR6oVaoNrGBoISnM7GqqCtlR+b9z mc1Q== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1754349358; x=1754954158; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:x-gm-message-state:from:to:cc:subject:date:message-id:reply-to; bh=rATS4q8eQgXRIMxeVI6i3f0N8VShHmldfKu+DIDZEpI=; b=Zc2/qd+7uiJtL8D4ZJU89oFdZBlgulSux8kHdlSPXD4LdgQ9/bpubEOOlzu270MtNX uyAlbYemsdD8FgcGHMmkarBkAWJ7vJ5cdHu8t40Bw933ICl/YNdu3s/IkZ3XZzlGd9R6 jktJUm1EaGMHck2z+aM7EKG3U/mzrhutlIH1j6onJzZ5xB2NbcsEZ3XRK4pPKYX0Wr8u iaEq3QW6amYY2XUwCZK07zWLxmSuVXXrw/y186O6A5yhP3XeoZZg/hH8pbQ+fAxyeYH4 Rc5fjisJW+dFkMCUMAHYNMpkfy292QRK/8dVGQZgvgDnCWQmAFOmGRPc8G18FUefFFuq S1xw== X-Forwarded-Encrypted: i=1; AJvYcCX5dvgbVWRZKafkIoZMRqpcYfH99qdknND+bBgAqX1lwkeF+zoGE01yJcGBxccrCDndZ/BNDqoaba6pcrU=@vger.kernel.org X-Gm-Message-State: AOJu0Yzhr5hF2MpWF2a6owXn5WAnGb0RxMD21mYjl0TX5jWuLMrwcNE5 1Ksmp7uGT0UjFMp1v2Ay0rHuJ6f6pWwkbXLO74sGglV2CAXNaEnGKFAksOw2iz1Xnk1BHmO5Lwg VZBoN2g== X-Google-Smtp-Source: AGHT+IEuSOf2jMTxPxDkqYSl5LW5Zcze6ynU2QOTKW6hn3gYFMFUvn0B/NfoN2M6d8RkK8dOvKSIuLnzjmM= X-Received: from pjl8.prod.google.com ([2002:a17:90b:2f88:b0:31e:d618:a29c]) (user=surenb job=prod-delivery.src-stubby-dispatcher) by 2002:a17:90b:4a88:b0:31f:867:d6b4 with SMTP id 98e67ed59e1d1-321161f0711mr16728967a91.10.1754349358011; Mon, 04 Aug 2025 16:15:58 -0700 (PDT) Date: Mon, 4 Aug 2025 16:15:49 -0700 In-Reply-To: <20250804231552.1217132-1-surenb@google.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: Mime-Version: 1.0 References: <20250804231552.1217132-1-surenb@google.com> X-Mailer: git-send-email 2.50.1.565.gc32cd1483b-goog Message-ID: <20250804231552.1217132-2-surenb@google.com> Subject: [PATCH v2 1/3] selftests/proc: test PROCMAP_QUERY ioctl while vma is concurrently modified From: Suren Baghdasaryan To: akpm@linux-foundation.org Cc: Liam.Howlett@oracle.com, lorenzo.stoakes@oracle.com, david@redhat.com, vbabka@suse.cz, peterx@redhat.com, jannh@google.com, hannes@cmpxchg.org, mhocko@kernel.org, paulmck@kernel.org, shuah@kernel.org, adobriyan@gmail.com, brauner@kernel.org, josef@toxicpanda.com, yebin10@huawei.com, linux@weissschuh.net, willy@infradead.org, osalvador@suse.de, andrii@kernel.org, ryan.roberts@arm.com, christophe.leroy@csgroup.eu, tjmercier@google.com, kaleshsingh@google.com, aha310510@gmail.com, linux-kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org, linux-mm@kvack.org, linux-kselftest@vger.kernel.org, surenb@google.com, SeongJae Park Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" Extend /proc/pid/maps tearing tests to verify PROCMAP_QUERY ioctl operation correctness while the vma is being concurrently modified. Signed-off-by: Suren Baghdasaryan Tested-by: SeongJae Park Acked-by: SeongJae Park --- tools/testing/selftests/proc/proc-maps-race.c | 65 +++++++++++++++++++ 1 file changed, 65 insertions(+) diff --git a/tools/testing/selftests/proc/proc-maps-race.c b/tools/testing/= selftests/proc/proc-maps-race.c index 66773685a047..d40854a07ec1 100644 --- a/tools/testing/selftests/proc/proc-maps-race.c +++ b/tools/testing/selftests/proc/proc-maps-race.c @@ -32,6 +32,8 @@ #include #include #include +#include +#include #include #include #include @@ -317,6 +319,25 @@ static bool capture_mod_pattern(FIXTURE_DATA(proc_maps= _race) *self, strcmp(restored_first_line->text, self->first_line.text) =3D=3D 0; } =20 +static bool query_addr_at(int maps_fd, void *addr, + unsigned long *vma_start, unsigned long *vma_end) +{ + struct procmap_query q; + + memset(&q, 0, sizeof(q)); + q.size =3D sizeof(q); + /* Find the VMA at the split address */ + q.query_addr =3D (unsigned long long)addr; + q.query_flags =3D 0; + if (ioctl(maps_fd, PROCMAP_QUERY, &q)) + return false; + + *vma_start =3D q.vma_start; + *vma_end =3D q.vma_end; + + return true; +} + static inline bool split_vma(FIXTURE_DATA(proc_maps_race) *self) { return mmap(self->mod_info->addr, self->page_size, self->mod_info->prot |= PROT_EXEC, @@ -559,6 +580,8 @@ TEST_F(proc_maps_race, test_maps_tearing_from_split) do { bool last_line_changed; bool first_line_changed; + unsigned long vma_start; + unsigned long vma_end; =20 ASSERT_TRUE(read_boundary_lines(self, &new_last_line, &new_first_line)); =20 @@ -595,6 +618,19 @@ TEST_F(proc_maps_race, test_maps_tearing_from_split) first_line_changed =3D strcmp(new_first_line.text, self->first_line.text= ) !=3D 0; ASSERT_EQ(last_line_changed, first_line_changed); =20 + /* Check if PROCMAP_QUERY ioclt() finds the right VMA */ + ASSERT_TRUE(query_addr_at(self->maps_fd, mod_info->addr + self->page_siz= e, + &vma_start, &vma_end)); + /* + * The vma at the split address can be either the same as + * original one (if read before the split) or the same as the + * first line in the second page (if read after the split). + */ + ASSERT_TRUE((vma_start =3D=3D self->last_line.start_addr && + vma_end =3D=3D self->last_line.end_addr) || + (vma_start =3D=3D split_first_line.start_addr && + vma_end =3D=3D split_first_line.end_addr)); + clock_gettime(CLOCK_MONOTONIC_COARSE, &end_ts); end_test_iteration(&end_ts, self->verbose); } while (end_ts.tv_sec - start_ts.tv_sec < self->duration_sec); @@ -636,6 +672,9 @@ TEST_F(proc_maps_race, test_maps_tearing_from_resize) clock_gettime(CLOCK_MONOTONIC_COARSE, &start_ts); start_test_loop(&start_ts, self->verbose); do { + unsigned long vma_start; + unsigned long vma_end; + ASSERT_TRUE(read_boundary_lines(self, &new_last_line, &new_first_line)); =20 /* Check if we read vmas after shrinking it */ @@ -662,6 +701,16 @@ TEST_F(proc_maps_race, test_maps_tearing_from_resize) "Expand result invalid", self)); } =20 + /* Check if PROCMAP_QUERY ioclt() finds the right VMA */ + ASSERT_TRUE(query_addr_at(self->maps_fd, mod_info->addr, &vma_start, &vm= a_end)); + /* + * The vma should stay at the same address and have either the + * original size of 3 pages or 1 page if read after shrinking. + */ + ASSERT_TRUE(vma_start =3D=3D self->last_line.start_addr && + (vma_end - vma_start =3D=3D self->page_size * 3 || + vma_end - vma_start =3D=3D self->page_size)); + clock_gettime(CLOCK_MONOTONIC_COARSE, &end_ts); end_test_iteration(&end_ts, self->verbose); } while (end_ts.tv_sec - start_ts.tv_sec < self->duration_sec); @@ -703,6 +752,9 @@ TEST_F(proc_maps_race, test_maps_tearing_from_remap) clock_gettime(CLOCK_MONOTONIC_COARSE, &start_ts); start_test_loop(&start_ts, self->verbose); do { + unsigned long vma_start; + unsigned long vma_end; + ASSERT_TRUE(read_boundary_lines(self, &new_last_line, &new_first_line)); =20 /* Check if we read vmas after remapping it */ @@ -729,6 +781,19 @@ TEST_F(proc_maps_race, test_maps_tearing_from_remap) "Remap restore result invalid", self)); } =20 + /* Check if PROCMAP_QUERY ioclt() finds the right VMA */ + ASSERT_TRUE(query_addr_at(self->maps_fd, mod_info->addr + self->page_siz= e, + &vma_start, &vma_end)); + /* + * The vma should either stay at the same address and have the + * original size of 3 pages or we should find the remapped vma + * at the remap destination address with size of 1 page. + */ + ASSERT_TRUE((vma_start =3D=3D self->last_line.start_addr && + vma_end - vma_start =3D=3D self->page_size * 3) || + (vma_start =3D=3D self->last_line.start_addr + self->page_size && + vma_end - vma_start =3D=3D self->page_size)); + clock_gettime(CLOCK_MONOTONIC_COARSE, &end_ts); end_test_iteration(&end_ts, self->verbose); } while (end_ts.tv_sec - start_ts.tv_sec < self->duration_sec); --=20 2.50.1.565.gc32cd1483b-goog From nobody Sun Oct 5 12:52:24 2025 Received: from mail-pl1-f201.google.com (mail-pl1-f201.google.com [209.85.214.201]) (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 EEBED1DE8B5 for ; Mon, 4 Aug 2025 23:16:00 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.214.201 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1754349362; cv=none; b=Zt53tiaNztAVLGde+XSVx1jSNfibMgFjzsShVslTK6ih7KWnTX5qc9G+z7bomHf8Hi00vUVb/HOQS9Lo4fVQgBs9KlT+OdSfl8/mdVTocOEActGM5VSU8WSmsyjhi1xJMczk7y28m5hhBK9mH5qfq+79Dw8dOQe8Pz0Q5lGtboo= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1754349362; c=relaxed/simple; bh=hbEq1Js8AlyKGQUt5xmaKCswYweUQKdmzuSbhBMwHKY=; h=Date:In-Reply-To:Mime-Version:References:Message-ID:Subject:From: To:Cc:Content-Type; b=CDM//kPgUtO1KbN9MaVQ5zK4XELiGLE9dqS8r+ZHOFlgY5n3ssQTqn3bpLYybCt//VCAz+PUnzRU9ucxrXEuUn6ogbh0VaTBf6e8KUOabu1EOoT6cbIAsSBbXwYGE5wEq7St6d2hHztd4eYfndLJX7okjnVIOHYokpE5CYS3kK0= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com; spf=pass smtp.mailfrom=flex--surenb.bounces.google.com; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b=BvsNY3s2; arc=none smtp.client-ip=209.85.214.201 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=flex--surenb.bounces.google.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b="BvsNY3s2" Received: by mail-pl1-f201.google.com with SMTP id d9443c01a7336-23fe26e5a33so59955725ad.3 for ; Mon, 04 Aug 2025 16:16:00 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20230601; t=1754349360; x=1754954160; darn=vger.kernel.org; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:from:to:cc:subject:date:message-id:reply-to; bh=7LLlVoqHDMFMIDh4+31o+ONV2fLHn7v9sHgdKTjexS4=; b=BvsNY3s2Jq837raVXdbyUux5M+QPo/2MU3rMcxdRFYcUSRBzoqNIZRD3VxwaMTKs3k CCX8ip2kzkSrlU5YOgqq/j73WbuFMimBvAtjkcdj9rSSHJl5mZjts1fULuShxK8rwmu9 WRTO+5mohOg6Ta0w9Ye4MOtlTtSQcZEbBk165XAsuSrC7iQeKzbyXffH+/vAbF4DPxPz KfWdjKWpEj9rhKdBX5ypKw9OYJCnamruwR+Vo3hSeroevZ51zcEVaWi5LWprmeiWH/wQ IaWEoPOYPgOIcPlARQouSoFDSrhJIjad9L17o/VHe2oDiZMkdEtbNinbewHfVDNHDu3c KZdw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1754349360; x=1754954160; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:x-gm-message-state:from:to:cc:subject:date:message-id:reply-to; bh=7LLlVoqHDMFMIDh4+31o+ONV2fLHn7v9sHgdKTjexS4=; b=Ebb0tDVNnSZXaXrvZHnHaTIrecYv5fcp2AIUWyfVVjRf75m8wWGRWtVvSnl6FpHlSs pRvcP2Iwu59AVrtotHVLe57UvqQKKM60w1HjxIOxcHwcYbSeMKzOk9ovCzju95jECyop FJNvmYo8aCM9sPQ6X/HVL8CZIbdW1fvz4coKQC12QMVvPrcoYZxdQbrcSicFQ3LDADbY rlIAo3EZbxzzpWxDqY+fwdMedAQsBoJZKzSCOv4OE/B6NmG9eG8tL6e/+RnZpDjTb4er KD7Tup0GjLBkcznc4jfyEqwtB7/wuno/1OlhW60rk+9N6syuGsmo+kPKkBJzO7EO3Ssu bzHg== X-Forwarded-Encrypted: i=1; AJvYcCUfD1ukWQ6VxQvEcBXa/+C32ZLLfoaLmFwyAuv/RkQxeq68wjtC5Y+1hel4PhLy2Ov7f1k3jBitgp3XjK8=@vger.kernel.org X-Gm-Message-State: AOJu0YzmT/raopznsihTGRx8k3KE7Z0Gxxz9LNshPVLODL8CD952f1Dp LeO1aBius8cmGqk6JE/eUyFCEZhrubArzfNON8+wAukmb98Ogx6+DtiYpcRGPVRFQ9EEE+76Yyv wtfSxrA== X-Google-Smtp-Source: AGHT+IF/QZzIgmBa6VwolABoKQDtTpPz3bEroyjsnVJNVCeHAjwJivrabr8q4l1Lq6BpaMaysvV3u9CBYtw= X-Received: from plcx13.prod.google.com ([2002:a17:903:cd:b0:240:1be2:19ee]) (user=surenb job=prod-delivery.src-stubby-dispatcher) by 2002:a17:902:cec4:b0:240:48f4:40d5 with SMTP id d9443c01a7336-24247023e9cmr137793735ad.39.1754349360090; Mon, 04 Aug 2025 16:16:00 -0700 (PDT) Date: Mon, 4 Aug 2025 16:15:50 -0700 In-Reply-To: <20250804231552.1217132-1-surenb@google.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: Mime-Version: 1.0 References: <20250804231552.1217132-1-surenb@google.com> X-Mailer: git-send-email 2.50.1.565.gc32cd1483b-goog Message-ID: <20250804231552.1217132-3-surenb@google.com> Subject: [PATCH v2 2/3] fs/proc/task_mmu: factor out proc_maps_private fields used by PROCMAP_QUERY From: Suren Baghdasaryan To: akpm@linux-foundation.org Cc: Liam.Howlett@oracle.com, lorenzo.stoakes@oracle.com, david@redhat.com, vbabka@suse.cz, peterx@redhat.com, jannh@google.com, hannes@cmpxchg.org, mhocko@kernel.org, paulmck@kernel.org, shuah@kernel.org, adobriyan@gmail.com, brauner@kernel.org, josef@toxicpanda.com, yebin10@huawei.com, linux@weissschuh.net, willy@infradead.org, osalvador@suse.de, andrii@kernel.org, ryan.roberts@arm.com, christophe.leroy@csgroup.eu, tjmercier@google.com, kaleshsingh@google.com, aha310510@gmail.com, linux-kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org, linux-mm@kvack.org, linux-kselftest@vger.kernel.org, surenb@google.com Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" Refactor struct proc_maps_private so that the fields used by PROCMAP_QUERY ioctl are moved into a separate structure. In the next patch this allows ioctl to reuse some of the functions used for reading /proc/pid/maps without using file->private_data. This prevents concurrent modification of file->private_data members by ioctl and /proc/pid/maps readers. The change is pure code refactoring and has no functional changes. Signed-off-by: Suren Baghdasaryan --- fs/proc/internal.h | 15 ++++++---- fs/proc/task_mmu.c | 70 ++++++++++++++++++++++---------------------- fs/proc/task_nommu.c | 14 ++++----- 3 files changed, 52 insertions(+), 47 deletions(-) diff --git a/fs/proc/internal.h b/fs/proc/internal.h index 7c235451c5ea..663dc0510315 100644 --- a/fs/proc/internal.h +++ b/fs/proc/internal.h @@ -379,16 +379,21 @@ extern void proc_self_init(void); * task_[no]mmu.c */ struct mem_size_stats; -struct proc_maps_private { - struct inode *inode; - struct task_struct *task; + +struct proc_maps_locking_ctx { struct mm_struct *mm; - struct vma_iterator iter; - loff_t last_pos; #ifdef CONFIG_PER_VMA_LOCK bool mmap_locked; struct vm_area_struct *locked_vma; #endif +}; + +struct proc_maps_private { + struct inode *inode; + struct task_struct *task; + struct vma_iterator iter; + loff_t last_pos; + struct proc_maps_locking_ctx lock_ctx; #ifdef CONFIG_NUMA struct mempolicy *task_mempolicy; #endif diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c index 3d6d8a9f13fc..843577aa7a32 100644 --- a/fs/proc/task_mmu.c +++ b/fs/proc/task_mmu.c @@ -132,11 +132,11 @@ static void release_task_mempolicy(struct proc_maps_p= rivate *priv) =20 #ifdef CONFIG_PER_VMA_LOCK =20 -static void unlock_vma(struct proc_maps_private *priv) +static void unlock_vma(struct proc_maps_locking_ctx *lock_ctx) { - if (priv->locked_vma) { - vma_end_read(priv->locked_vma); - priv->locked_vma =3D NULL; + if (lock_ctx->locked_vma) { + vma_end_read(lock_ctx->locked_vma); + lock_ctx->locked_vma =3D NULL; } } =20 @@ -151,14 +151,14 @@ static inline bool lock_vma_range(struct seq_file *m, * walking the vma tree under rcu read protection. */ if (m->op !=3D &proc_pid_maps_op) { - if (mmap_read_lock_killable(priv->mm)) + if (mmap_read_lock_killable(priv->lock_ctx.mm)) return false; =20 - priv->mmap_locked =3D true; + priv->lock_ctx.mmap_locked =3D true; } else { rcu_read_lock(); - priv->locked_vma =3D NULL; - priv->mmap_locked =3D false; + priv->lock_ctx.locked_vma =3D NULL; + priv->lock_ctx.mmap_locked =3D false; } =20 return true; @@ -166,10 +166,10 @@ static inline bool lock_vma_range(struct seq_file *m, =20 static inline void unlock_vma_range(struct proc_maps_private *priv) { - if (priv->mmap_locked) { - mmap_read_unlock(priv->mm); + if (priv->lock_ctx.mmap_locked) { + mmap_read_unlock(priv->lock_ctx.mm); } else { - unlock_vma(priv); + unlock_vma(&priv->lock_ctx); rcu_read_unlock(); } } @@ -179,13 +179,13 @@ static struct vm_area_struct *get_next_vma(struct pro= c_maps_private *priv, { struct vm_area_struct *vma; =20 - if (priv->mmap_locked) + if (priv->lock_ctx.mmap_locked) return vma_next(&priv->iter); =20 - unlock_vma(priv); - vma =3D lock_next_vma(priv->mm, &priv->iter, last_pos); + unlock_vma(&priv->lock_ctx); + vma =3D lock_next_vma(priv->lock_ctx.mm, &priv->iter, last_pos); if (!IS_ERR_OR_NULL(vma)) - priv->locked_vma =3D vma; + priv->lock_ctx.locked_vma =3D vma; =20 return vma; } @@ -193,14 +193,14 @@ static struct vm_area_struct *get_next_vma(struct pro= c_maps_private *priv, static inline bool fallback_to_mmap_lock(struct proc_maps_private *priv, loff_t pos) { - if (priv->mmap_locked) + if (priv->lock_ctx.mmap_locked) return false; =20 rcu_read_unlock(); - mmap_read_lock(priv->mm); + mmap_read_lock(priv->lock_ctx.mm); /* Reinitialize the iterator after taking mmap_lock */ vma_iter_set(&priv->iter, pos); - priv->mmap_locked =3D true; + priv->lock_ctx.mmap_locked =3D true; =20 return true; } @@ -210,12 +210,12 @@ static inline bool fallback_to_mmap_lock(struct proc_= maps_private *priv, static inline bool lock_vma_range(struct seq_file *m, struct proc_maps_private *priv) { - return mmap_read_lock_killable(priv->mm) =3D=3D 0; + return mmap_read_lock_killable(priv->lock_ctx.mm) =3D=3D 0; } =20 static inline void unlock_vma_range(struct proc_maps_private *priv) { - mmap_read_unlock(priv->mm); + mmap_read_unlock(priv->lock_ctx.mm); } =20 static struct vm_area_struct *get_next_vma(struct proc_maps_private *priv, @@ -258,7 +258,7 @@ static struct vm_area_struct *proc_get_vma(struct seq_f= ile *m, loff_t *ppos) *ppos =3D vma->vm_end; } else { *ppos =3D SENTINEL_VMA_GATE; - vma =3D get_gate_vma(priv->mm); + vma =3D get_gate_vma(priv->lock_ctx.mm); } =20 return vma; @@ -278,7 +278,7 @@ static void *m_start(struct seq_file *m, loff_t *ppos) if (!priv->task) return ERR_PTR(-ESRCH); =20 - mm =3D priv->mm; + mm =3D priv->lock_ctx.mm; if (!mm || !mmget_not_zero(mm)) { put_task_struct(priv->task); priv->task =3D NULL; @@ -318,7 +318,7 @@ static void *m_next(struct seq_file *m, void *v, loff_t= *ppos) static void m_stop(struct seq_file *m, void *v) { struct proc_maps_private *priv =3D m->private; - struct mm_struct *mm =3D priv->mm; + struct mm_struct *mm =3D priv->lock_ctx.mm; =20 if (!priv->task) return; @@ -339,9 +339,9 @@ static int proc_maps_open(struct inode *inode, struct f= ile *file, return -ENOMEM; =20 priv->inode =3D inode; - priv->mm =3D proc_mem_open(inode, PTRACE_MODE_READ); - if (IS_ERR_OR_NULL(priv->mm)) { - int err =3D priv->mm ? PTR_ERR(priv->mm) : -ESRCH; + priv->lock_ctx.mm =3D proc_mem_open(inode, PTRACE_MODE_READ); + if (IS_ERR_OR_NULL(priv->lock_ctx.mm)) { + int err =3D priv->lock_ctx.mm ? PTR_ERR(priv->lock_ctx.mm) : -ESRCH; =20 seq_release_private(inode, file); return err; @@ -355,8 +355,8 @@ static int proc_map_release(struct inode *inode, struct= file *file) struct seq_file *seq =3D file->private_data; struct proc_maps_private *priv =3D seq->private; =20 - if (priv->mm) - mmdrop(priv->mm); + if (priv->lock_ctx.mm) + mmdrop(priv->lock_ctx.mm); =20 return seq_release_private(inode, file); } @@ -610,7 +610,7 @@ static int do_procmap_query(struct proc_maps_private *p= riv, void __user *uarg) if (!!karg.build_id_size !=3D !!karg.build_id_addr) return -EINVAL; =20 - mm =3D priv->mm; + mm =3D priv->lock_ctx.mm; if (!mm || !mmget_not_zero(mm)) return -ESRCH; =20 @@ -1307,7 +1307,7 @@ static int show_smaps_rollup(struct seq_file *m, void= *v) { struct proc_maps_private *priv =3D m->private; struct mem_size_stats mss =3D {}; - struct mm_struct *mm =3D priv->mm; + struct mm_struct *mm =3D priv->lock_ctx.mm; struct vm_area_struct *vma; unsigned long vma_start =3D 0, last_vma_end =3D 0; int ret =3D 0; @@ -1452,9 +1452,9 @@ static int smaps_rollup_open(struct inode *inode, str= uct file *file) goto out_free; =20 priv->inode =3D inode; - priv->mm =3D proc_mem_open(inode, PTRACE_MODE_READ); - if (IS_ERR_OR_NULL(priv->mm)) { - ret =3D priv->mm ? PTR_ERR(priv->mm) : -ESRCH; + priv->lock_ctx.mm =3D proc_mem_open(inode, PTRACE_MODE_READ); + if (IS_ERR_OR_NULL(priv->lock_ctx.mm)) { + ret =3D priv->lock_ctx.mm ? PTR_ERR(priv->lock_ctx.mm) : -ESRCH; =20 single_release(inode, file); goto out_free; @@ -1472,8 +1472,8 @@ static int smaps_rollup_release(struct inode *inode, = struct file *file) struct seq_file *seq =3D file->private_data; struct proc_maps_private *priv =3D seq->private; =20 - if (priv->mm) - mmdrop(priv->mm); + if (priv->lock_ctx.mm) + mmdrop(priv->lock_ctx.mm); =20 kfree(priv); return single_release(inode, file); diff --git a/fs/proc/task_nommu.c b/fs/proc/task_nommu.c index 59bfd61d653a..d362919f4f68 100644 --- a/fs/proc/task_nommu.c +++ b/fs/proc/task_nommu.c @@ -204,7 +204,7 @@ static void *m_start(struct seq_file *m, loff_t *ppos) if (!priv->task) return ERR_PTR(-ESRCH); =20 - mm =3D priv->mm; + mm =3D priv->lock_ctx.mm; if (!mm || !mmget_not_zero(mm)) { put_task_struct(priv->task); priv->task =3D NULL; @@ -226,7 +226,7 @@ static void *m_start(struct seq_file *m, loff_t *ppos) static void m_stop(struct seq_file *m, void *v) { struct proc_maps_private *priv =3D m->private; - struct mm_struct *mm =3D priv->mm; + struct mm_struct *mm =3D priv->lock_ctx.mm; =20 if (!priv->task) return; @@ -259,9 +259,9 @@ static int maps_open(struct inode *inode, struct file *= file, return -ENOMEM; =20 priv->inode =3D inode; - priv->mm =3D proc_mem_open(inode, PTRACE_MODE_READ); - if (IS_ERR_OR_NULL(priv->mm)) { - int err =3D priv->mm ? PTR_ERR(priv->mm) : -ESRCH; + priv->lock_ctx.mm =3D proc_mem_open(inode, PTRACE_MODE_READ); + if (IS_ERR_OR_NULL(priv->lock_ctx.mm)) { + int err =3D priv->lock_ctx.mm ? PTR_ERR(priv->lock_ctx.mm) : -ESRCH; =20 seq_release_private(inode, file); return err; @@ -276,8 +276,8 @@ static int map_release(struct inode *inode, struct file= *file) struct seq_file *seq =3D file->private_data; struct proc_maps_private *priv =3D seq->private; =20 - if (priv->mm) - mmdrop(priv->mm); + if (priv->lock_ctx.mm) + mmdrop(priv->lock_ctx.mm); =20 return seq_release_private(inode, file); } --=20 2.50.1.565.gc32cd1483b-goog From nobody Sun Oct 5 12:52:24 2025 Received: from mail-pf1-f202.google.com (mail-pf1-f202.google.com [209.85.210.202]) (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 02AD5252286 for ; Mon, 4 Aug 2025 23:16:02 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.210.202 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1754349364; cv=none; b=UhxeWvvZlyKCEB9LaJbFNEocbtfUR0FYbi5z7qZ6CiN2KHIzW/HcskhQ9cbMUSE9N5s2wCQwNViI293wIBEaQ9qU+ShFI8W7fMLHCO42r4MmZh2rUBVIer+UrIPO+VonAslIh0VyaD5g95qaCc9X7oDPgw+t6nyjXf4uO/rNFZs= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1754349364; c=relaxed/simple; bh=BdiowjuFZsdquH1v1qJ0l1K5TohHeQmAGtriPnbOYKY=; h=Date:In-Reply-To:Mime-Version:References:Message-ID:Subject:From: To:Cc:Content-Type; b=lFU+bzWyXe0BEdvL+75xPKT/yc++nqCNLwIFXNiYfLzC5T2wdSuNn4bwyz7hfM2Kf6gB23O0hZT+opvfqZ8Rr9VU2Lstax2d+3b8cLGSJbLFHal8UFjYeFqxp6Yx8gDk/T5syBdbClWxdWs9KO/MphlACCXYshtBLFqR/u8vL6A= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=google.com; spf=pass smtp.mailfrom=flex--surenb.bounces.google.com; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b=E5VDgKCu; arc=none smtp.client-ip=209.85.210.202 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=flex--surenb.bounces.google.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=google.com header.i=@google.com header.b="E5VDgKCu" Received: by mail-pf1-f202.google.com with SMTP id d2e1a72fcca58-74d15d90cdbso3486169b3a.0 for ; Mon, 04 Aug 2025 16:16:02 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20230601; t=1754349362; x=1754954162; darn=vger.kernel.org; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:from:to:cc:subject:date:message-id:reply-to; bh=Nn/QlXxKr04lsk0BQprsWZ8fvAafoan6Kc4Ifd428WQ=; b=E5VDgKCuzMlzcOS0yRUa7w4xr22v9RyPFr+ZvBv0mDMtfP959jmEUe7cZNC54RaYsZ DkEByUH8FOmw4BShu7TtMaWFxy6vpDeuSxPw4BWw5D9NTLFMCxNCgyBEvLyPwpBtvxm6 7mkUqE9dXu0D7rUibyPddAhMu/1VNJloSmhzU2d9PvhQ/30wYFd/cZ6mRf+5bkDq+e54 m6dQXBC4eDCb0SIWZ3byJGB5DfnLg62SCHIxfVVa35SaoiLjcLiSF7750y4IlrxT0Rq9 dM3FpbtXQDeZPT2R7EyyQ7TkxlErxXODBQ5eZPgJWPsDS19/N8w2AoSfRcZAn2P+5SpL +H7g== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1754349362; x=1754954162; h=cc:to:from:subject:message-id:references:mime-version:in-reply-to :date:x-gm-message-state:from:to:cc:subject:date:message-id:reply-to; bh=Nn/QlXxKr04lsk0BQprsWZ8fvAafoan6Kc4Ifd428WQ=; b=q+1sR9h7wS8mQiZl0KxYD7TbL/THBZ1uGwLXjGpGjgbv8R4x/RxqSQoa4M/fghy2Ma 0GKViTvpeZxEAAm52EQIuoDqqo6NtjmwaLlBzaUN1HBIz2ml47pXTHtn2tcbqYFULsSV WaHPeetfN53HSC2S0LxiyQZnV+UGuqWF/k1xmKbqlHxqoupLI/+Z6vceQhLcOn3+iLOR P+tTwkoXUyOSn4m7GGNqHuEFty6OYXlwZAkZdfHwdOWu9Upv+ZJv+GjeVA4h2V89+ulV R+aJ+Q3SeJGVBiuiR82aFP69WGZ7y+BYHGDV8DgkDXcd84fgkjz+QVn+n8aO6fajXzEJ B8Vg== X-Forwarded-Encrypted: i=1; AJvYcCUBQBB6qxIp73T5v5huu1UIy/mwtyojV4OCE+PzcpyNSPCtQdO6M8jw6LMhOkD+1Ka2zmop0E5R4BIw/Ek=@vger.kernel.org X-Gm-Message-State: AOJu0YyKBGP4RJuD45G8LuVfor1C5Vu+NHKKineAiiLLqxFjQtjGMpfj ahXMbZ+7ievQcR5Szmi+pRWqhHKsPseN33mou/B13/JhMfbeXs+vAA6wKhL8Pd6LSFcI2sEBirb msfqOVw== X-Google-Smtp-Source: AGHT+IHSKj2r68iHuJ6mz0Uqid6yqFPSiOF1csXgv0QqUAcL3Sg9DHhAyqp/y83a4PMGePmffR52Td+bl/E= X-Received: from pgnh8.prod.google.com ([2002:a63:3848:0:b0:b42:14da:59ac]) (user=surenb job=prod-delivery.src-stubby-dispatcher) by 2002:a05:6a20:244f:b0:240:d12:775c with SMTP id adf61e73a8af0-2400d12a143mr7876300637.36.1754349362131; Mon, 04 Aug 2025 16:16:02 -0700 (PDT) Date: Mon, 4 Aug 2025 16:15:51 -0700 In-Reply-To: <20250804231552.1217132-1-surenb@google.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: Mime-Version: 1.0 References: <20250804231552.1217132-1-surenb@google.com> X-Mailer: git-send-email 2.50.1.565.gc32cd1483b-goog Message-ID: <20250804231552.1217132-4-surenb@google.com> Subject: [PATCH v2 3/3] fs/proc/task_mmu: execute PROCMAP_QUERY ioctl under per-vma locks From: Suren Baghdasaryan To: akpm@linux-foundation.org Cc: Liam.Howlett@oracle.com, lorenzo.stoakes@oracle.com, david@redhat.com, vbabka@suse.cz, peterx@redhat.com, jannh@google.com, hannes@cmpxchg.org, mhocko@kernel.org, paulmck@kernel.org, shuah@kernel.org, adobriyan@gmail.com, brauner@kernel.org, josef@toxicpanda.com, yebin10@huawei.com, linux@weissschuh.net, willy@infradead.org, osalvador@suse.de, andrii@kernel.org, ryan.roberts@arm.com, christophe.leroy@csgroup.eu, tjmercier@google.com, kaleshsingh@google.com, aha310510@gmail.com, linux-kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org, linux-mm@kvack.org, linux-kselftest@vger.kernel.org, surenb@google.com Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" Utilize per-vma locks to stabilize vma after lookup without taking mmap_lock during PROCMAP_QUERY ioctl execution. If vma lock is contended, we fall back to mmap_lock but take it only momentarily to lock the vma and release the mmap_lock. In a very unlikely case of vm_refcnt overflow, this fall back path will fail and ioctl is done under mmap_lock protection. This change is designed to reduce mmap_lock contention and prevent PROCMAP_QUERY ioctl calls from blocking address space updates. Signed-off-by: Suren Baghdasaryan --- fs/proc/task_mmu.c | 81 +++++++++++++++++++++++++++++++++++++--------- 1 file changed, 65 insertions(+), 16 deletions(-) diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c index 843577aa7a32..1d06ecdbef6f 100644 --- a/fs/proc/task_mmu.c +++ b/fs/proc/task_mmu.c @@ -517,28 +517,78 @@ static int pid_maps_open(struct inode *inode, struct = file *file) PROCMAP_QUERY_VMA_FLAGS \ ) =20 -static int query_vma_setup(struct mm_struct *mm) +#ifdef CONFIG_PER_VMA_LOCK + +static int query_vma_setup(struct proc_maps_locking_ctx *lock_ctx) { - return mmap_read_lock_killable(mm); + lock_ctx->locked_vma =3D NULL; + lock_ctx->mmap_locked =3D false; + + return 0; } =20 -static void query_vma_teardown(struct mm_struct *mm, struct vm_area_struct= *vma) +static void query_vma_teardown(struct proc_maps_locking_ctx *lock_ctx) { - mmap_read_unlock(mm); + if (lock_ctx->mmap_locked) + mmap_read_unlock(lock_ctx->mm); + else + unlock_vma(lock_ctx); } =20 -static struct vm_area_struct *query_vma_find_by_addr(struct mm_struct *mm,= unsigned long addr) +static struct vm_area_struct *query_vma_find_by_addr(struct proc_maps_lock= ing_ctx *lock_ctx, + unsigned long addr) { - return find_vma(mm, addr); + struct vm_area_struct *vma; + struct vma_iterator vmi; + + unlock_vma(lock_ctx); + rcu_read_lock(); + vma_iter_init(&vmi, lock_ctx->mm, addr); + vma =3D lock_next_vma(lock_ctx->mm, &vmi, addr); + rcu_read_unlock(); + + if (!IS_ERR_OR_NULL(vma)) { + lock_ctx->locked_vma =3D vma; + } else if (PTR_ERR(vma) =3D=3D -EAGAIN) { + /* Fallback to mmap_lock on vma->vm_refcnt overflow */ + mmap_read_lock(lock_ctx->mm); + vma =3D find_vma(lock_ctx->mm, addr); + lock_ctx->mmap_locked =3D true; + } + + return vma; } =20 -static struct vm_area_struct *query_matching_vma(struct mm_struct *mm, +#else /* CONFIG_PER_VMA_LOCK */ + +static int query_vma_setup(struct proc_maps_locking_ctx *lock_ctx) +{ + return mmap_read_lock_killable(lock_ctx->mm); +} + +static void query_vma_teardown(struct proc_maps_locking_ctx *lock_ctx) +{ + mmap_read_unlock(lock_ctx->mm); +} + +static struct vm_area_struct *query_vma_find_by_addr(struct proc_maps_lock= ing_ctx *lock_ctx, + unsigned long addr) +{ + return find_vma(lock_ctx->mm, addr); +} + +#endif /* CONFIG_PER_VMA_LOCK */ + +static struct vm_area_struct *query_matching_vma(struct proc_maps_locking_= ctx *lock_ctx, unsigned long addr, u32 flags) { struct vm_area_struct *vma; =20 next_vma: - vma =3D query_vma_find_by_addr(mm, addr); + vma =3D query_vma_find_by_addr(lock_ctx, addr); + if (IS_ERR(vma)) + return vma; + if (!vma) goto no_vma; =20 @@ -579,11 +629,11 @@ static struct vm_area_struct *query_matching_vma(stru= ct mm_struct *mm, return ERR_PTR(-ENOENT); } =20 -static int do_procmap_query(struct proc_maps_private *priv, void __user *u= arg) +static int do_procmap_query(struct mm_struct *mm, void __user *uarg) { + struct proc_maps_locking_ctx lock_ctx =3D { .mm =3D mm }; struct procmap_query karg; struct vm_area_struct *vma; - struct mm_struct *mm; const char *name =3D NULL; char build_id_buf[BUILD_ID_SIZE_MAX], *name_buf =3D NULL; __u64 usize; @@ -610,17 +660,16 @@ static int do_procmap_query(struct proc_maps_private = *priv, void __user *uarg) if (!!karg.build_id_size !=3D !!karg.build_id_addr) return -EINVAL; =20 - mm =3D priv->lock_ctx.mm; if (!mm || !mmget_not_zero(mm)) return -ESRCH; =20 - err =3D query_vma_setup(mm); + err =3D query_vma_setup(&lock_ctx); if (err) { mmput(mm); return err; } =20 - vma =3D query_matching_vma(mm, karg.query_addr, karg.query_flags); + vma =3D query_matching_vma(&lock_ctx, karg.query_addr, karg.query_flags); if (IS_ERR(vma)) { err =3D PTR_ERR(vma); vma =3D NULL; @@ -705,7 +754,7 @@ static int do_procmap_query(struct proc_maps_private *p= riv, void __user *uarg) } =20 /* unlock vma or mmap_lock, and put mm_struct before copying data to user= */ - query_vma_teardown(mm, vma); + query_vma_teardown(&lock_ctx); mmput(mm); =20 if (karg.vma_name_size && copy_to_user(u64_to_user_ptr(karg.vma_name_addr= ), @@ -725,7 +774,7 @@ static int do_procmap_query(struct proc_maps_private *p= riv, void __user *uarg) return 0; =20 out: - query_vma_teardown(mm, vma); + query_vma_teardown(&lock_ctx); mmput(mm); kfree(name_buf); return err; @@ -738,7 +787,7 @@ static long procfs_procmap_ioctl(struct file *file, uns= igned int cmd, unsigned l =20 switch (cmd) { case PROCMAP_QUERY: - return do_procmap_query(priv, (void __user *)arg); + return do_procmap_query(priv->lock_ctx.mm, (void __user *)arg); default: return -ENOIOCTLCMD; } --=20 2.50.1.565.gc32cd1483b-goog