Linux-stable-mirror April 2025

linux-stable-mirror@lists.linaro.org

562 participants
1507 discussions

[PATCH 6.1.y] smb: client: Fix use-after-free of network namespace.

by Xiangyu Chen

From: Kuniyuki Iwashima <kuniyu(a)amazon.com> commit ef7134c7fc48e1441b398e55a862232868a6f0a7 upstream. Recently, we got a customer report that CIFS triggers oops while reconnecting to a server. [0] The workload runs on Kubernetes, and some pods mount CIFS servers in non-root network namespaces. The problem rarely happened, but it was always while the pod was dying. The root cause is wrong reference counting for network namespace. CIFS uses kernel sockets, which do not hold refcnt of the netns that the socket belongs to. That means CIFS must ensure the socket is always freed before its netns; otherwise, use-after-free happens. The repro steps are roughly: 1. mount CIFS in a non-root netns 2. drop packets from the netns 3. destroy the netns 4. unmount CIFS We can reproduce the issue quickly with the script [1] below and see the splat [2] if CONFIG_NET_NS_REFCNT_TRACKER is enabled. When the socket is TCP, it is hard to guarantee the netns lifetime without holding refcnt due to async timers. Let's hold netns refcnt for each socket as done for SMC in commit 9744d2bf1976 ("smc: Fix use-after-free in tcp_write_timer_handler()."). Note that we need to move put_net() from cifs_put_tcp_session() to clean_demultiplex_info(); otherwise, __sock_create() still could touch a freed netns while cifsd tries to reconnect from cifs_demultiplex_thread(). Also, maybe_get_net() cannot be put just before __sock_create() because the code is not under RCU and there is a small chance that the same address happened to be reallocated to another netns. [0]: CIFS: VFS: \\XXXXXXXXXXX has not responded in 15 seconds. Reconnecting... CIFS: Serverclose failed 4 times, giving up Unable to handle kernel paging request at virtual address 14de99e461f84a07 Mem abort info: ESR = 0x0000000096000004 EC = 0x25: DABT (current EL), IL = 32 bits SET = 0, FnV = 0 EA = 0, S1PTW = 0 FSC = 0x04: level 0 translation fault Data abort info: ISV = 0, ISS = 0x00000004 CM = 0, WnR = 0 [14de99e461f84a07] address between user and kernel address ranges Internal error: Oops: 0000000096000004 [#1] SMP Modules linked in: cls_bpf sch_ingress nls_utf8 cifs cifs_arc4 cifs_md4 dns_resolver tcp_diag inet_diag veth xt_state xt_connmark nf_conntrack_netlink xt_nat xt_statistic xt_MASQUERADE xt_mark xt_addrtype ipt_REJECT nf_reject_ipv4 nft_chain_nat nf_nat xt_conntrack nf_conntrack nf_defrag_ipv6 nf_defrag_ipv4 xt_comment nft_compat nf_tables nfnetlink overlay nls_ascii nls_cp437 sunrpc vfat fat aes_ce_blk aes_ce_cipher ghash_ce sm4_ce_cipher sm4 sm3_ce sm3 sha3_ce sha512_ce sha512_arm64 sha1_ce ena button sch_fq_codel loop fuse configfs dmi_sysfs sha2_ce sha256_arm64 dm_mirror dm_region_hash dm_log dm_mod dax efivarfs CPU: 5 PID: 2690970 Comm: cifsd Not tainted 6.1.103-109.184.amzn2023.aarch64 #1 Hardware name: Amazon EC2 r7g.4xlarge/, BIOS 1.0 11/1/2018 pstate: 00400005 (nzcv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--) pc : fib_rules_lookup+0x44/0x238 lr : __fib_lookup+0x64/0xbc sp : ffff8000265db790 x29: ffff8000265db790 x28: 0000000000000000 x27: 000000000000bd01 x26: 0000000000000000 x25: ffff000b4baf8000 x24: ffff00047b5e4580 x23: ffff8000265db7e0 x22: 0000000000000000 x21: ffff00047b5e4500 x20: ffff0010e3f694f8 x19: 14de99e461f849f7 x18: 0000000000000000 x17: 0000000000000000 x16: 0000000000000000 x15: 0000000000000000 x14: 0000000000000000 x13: 0000000000000000 x12: 3f92800abd010002 x11: 0000000000000001 x10: ffff0010e3f69420 x9 : ffff800008a6f294 x8 : 0000000000000000 x7 : 0000000000000006 x6 : 0000000000000000 x5 : 0000000000000001 x4 : ffff001924354280 x3 : ffff8000265db7e0 x2 : 0000000000000000 x1 : ffff0010e3f694f8 x0 : ffff00047b5e4500 Call trace: fib_rules_lookup+0x44/0x238 __fib_lookup+0x64/0xbc ip_route_output_key_hash_rcu+0x2c4/0x398 ip_route_output_key_hash+0x60/0x8c tcp_v4_connect+0x290/0x488 __inet_stream_connect+0x108/0x3d0 inet_stream_connect+0x50/0x78 kernel_connect+0x6c/0xac generic_ip_connect+0x10c/0x6c8 [cifs] __reconnect_target_unlocked+0xa0/0x214 [cifs] reconnect_dfs_server+0x144/0x460 [cifs] cifs_reconnect+0x88/0x148 [cifs] cifs_readv_from_socket+0x230/0x430 [cifs] cifs_read_from_socket+0x74/0xa8 [cifs] cifs_demultiplex_thread+0xf8/0x704 [cifs] kthread+0xd0/0xd4 Code: aa0003f8 f8480f13 eb18027f 540006c0 (b9401264) [1]: CIFS_CRED="/root/cred.cifs" CIFS_USER="Administrator" CIFS_PASS="Password" CIFS_IP="X.X.X.X" CIFS_PATH="//${CIFS_IP}/Users/Administrator/Desktop/CIFS_TEST" CIFS_MNT="/mnt/smb" DEV="enp0s3" cat <<EOF > ${CIFS_CRED} username=${CIFS_USER} password=${CIFS_PASS} domain=EXAMPLE.COM EOF unshare -n bash -c " mkdir -p ${CIFS_MNT} ip netns attach root 1 ip link add eth0 type veth peer veth0 netns root ip link set eth0 up ip -n root link set veth0 up ip addr add 192.168.0.2/24 dev eth0 ip -n root addr add 192.168.0.1/24 dev veth0 ip route add default via 192.168.0.1 dev eth0 ip netns exec root sysctl net.ipv4.ip_forward=1 ip netns exec root iptables -t nat -A POSTROUTING -s 192.168.0.2 -o ${DEV} -j MASQUERADE mount -t cifs ${CIFS_PATH} ${CIFS_MNT} -o vers=3.0,sec=ntlmssp,credentials=${CIFS_CRED},rsize=65536,wsize=65536,cache=none,echo_interval=1 touch ${CIFS_MNT}/a.txt ip netns exec root iptables -t nat -D POSTROUTING -s 192.168.0.2 -o ${DEV} -j MASQUERADE " umount ${CIFS_MNT} [2]: ref_tracker: net notrefcnt@000000004bbc008d has 1/1 users at sk_alloc (./include/net/net_namespace.h:339 net/core/sock.c:2227) inet_create (net/ipv4/af_inet.c:326 net/ipv4/af_inet.c:252) __sock_create (net/socket.c:1576) generic_ip_connect (fs/smb/client/connect.c:3075) cifs_get_tcp_session.part.0 (fs/smb/client/connect.c:3160 fs/smb/client/connect.c:1798) cifs_mount_get_session (fs/smb/client/trace.h:959 fs/smb/client/connect.c:3366) dfs_mount_share (fs/smb/client/dfs.c:63 fs/smb/client/dfs.c:285) cifs_mount (fs/smb/client/connect.c:3622) cifs_smb3_do_mount (fs/smb/client/cifsfs.c:949) smb3_get_tree (fs/smb/client/fs_context.c:784 fs/smb/client/fs_context.c:802 fs/smb/client/fs_context.c:794) vfs_get_tree (fs/super.c:1800) path_mount (fs/namespace.c:3508 fs/namespace.c:3834) __x64_sys_mount (fs/namespace.c:3848 fs/namespace.c:4057 fs/namespace.c:4034 fs/namespace.c:4034) do_syscall_64 (arch/x86/entry/common.c:52 arch/x86/entry/common.c:83) entry_SYSCALL_64_after_hwframe (arch/x86/entry/entry_64.S:130) Fixes: 26abe14379f8 ("net: Modify sk_alloc to not reference count the netns of kernel sockets.") Signed-off-by: Kuniyuki Iwashima <kuniyu(a)amazon.com> Acked-by: Tom Talpey <tom(a)talpey.com> Signed-off-by: Steve French <stfrench(a)microsoft.com> [ cherry-pick from amazon-linux amazon-6.1.y/mainline ] Link: https://github.com/amazonlinux/linux/commit/12bb8c57cead Signed-off-by: Xiangyu Chen <xiangyu.chen(a)windriver.com> Signed-off-by: He Zhe <zhe.he(a)windriver.com> --- Verified the build test. --- fs/smb/client/connect.c | 13 ++++++++++--- 1 file changed, 10 insertions(+), 3 deletions(-) diff --git a/fs/smb/client/connect.c b/fs/smb/client/connect.c index db30c4b8a221..13dcf973e795 100644 --- a/fs/smb/client/connect.c +++ b/fs/smb/client/connect.c @@ -1062,6 +1062,7 @@ static void clean_demultiplex_info(struct TCP_Server_Info *server) */ } + put_net(cifs_net_ns(server)); #ifdef CONFIG_CIFS_DFS_UPCALL kfree(server->origin_fullpath); kfree(server->leaf_fullpath); @@ -1656,8 +1657,6 @@ cifs_put_tcp_session(struct TCP_Server_Info *server, int from_reconnect) /* srv_count can never go negative */ WARN_ON(server->srv_count < 0); - put_net(cifs_net_ns(server)); - list_del_init(&server->tcp_ses_list); spin_unlock(&cifs_tcp_ses_lock); @@ -3031,7 +3030,10 @@ generic_ip_connect(struct TCP_Server_Info *server) } if (socket == NULL) { - rc = __sock_create(cifs_net_ns(server), sfamily, SOCK_STREAM, + struct net *net = cifs_net_ns(server); + struct sock *sk; + + rc = __sock_create(net, sfamily, SOCK_STREAM, IPPROTO_TCP, &socket, 1); if (rc < 0) { cifs_server_dbg(VFS, "Error %d creating socket\n", rc); @@ -3039,6 +3041,11 @@ generic_ip_connect(struct TCP_Server_Info *server) return rc; } + sk = socket->sk; + sk->sk_net_refcnt = 1; + get_net_track(net, &sk->ns_tracker, GFP_KERNEL); + sock_inuse_add(net, 1); + /* BB other socket options to set KEEPALIVE, NODELAY? */ cifs_dbg(FYI, "Socket created\n"); server->ssocket = socket; -- 2.34.1

9 months, 1 week

[PATCH 5.10/5.15/6.1] HID: mcp2221: Set driver data before I2C adapter add

by Mikhail Lobanov

From: Hamish Martin <hamish.martin(a)alliedtelesis.co.nz> commit f2d4a5834638bbc967371b9168c0b481519f7c5e upstream. The process of adding an I2C adapter can invoke I2C accesses on that new adapter (see i2c_detect()). Ensure we have set the adapter's driver data to avoid null pointer dereferences in the xfer functions during the adapter add. This has been noted in the past and the same fix proposed but not completed. See: https://lore.kernel.org/lkml/ef597e73-ed71-168e-52af-0d19b03734ac@vigem.de/ Signed-off-by: Hamish Martin <hamish.martin(a)alliedtelesis.co.nz> Signed-off-by: Jiri Kosina <jkosina(a)suse.cz> [Mikhail Lobanov: older kernels use i2c_add_adapter() instead of devm_i2c_add_adapter(). This problem manifests itself in stable kernel branches: https://syzkaller.appspot.com/bug?extid=40dc62a33c759a65ad44, https://syzkaller.appspot.com/bug?extid=24b122f8e84c3f9d9d1d] Signed-off-by: Mikhail Lobanov <m.lobanov(a)rosa.ru> --- drivers/hid/hid-mcp2221.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/hid/hid-mcp2221.c b/drivers/hid/hid-mcp2221.c index 560eeec4035a..b0dc0edc69c2 100644 --- a/drivers/hid/hid-mcp2221.c +++ b/drivers/hid/hid-mcp2221.c @@ -879,12 +879,12 @@ static int mcp2221_probe(struct hid_device *hdev, snprintf(mcp->adapter.name, sizeof(mcp->adapter.name), "MCP2221 usb-i2c bridge"); + i2c_set_adapdata(&mcp->adapter, mcp); ret = i2c_add_adapter(&mcp->adapter); if (ret) { hid_err(hdev, "can't add usb-i2c adapter: %d\n", ret); goto err_i2c; } - i2c_set_adapdata(&mcp->adapter, mcp); /* Setup GPIO chip */ mcp->gc = devm_kzalloc(&hdev->dev, sizeof(*mcp->gc), GFP_KERNEL); -- 2.47.2

9 months, 1 week

[PATCH] habanalabs: Add error handling for hl_mmu_get_hop_pte_phys_addr()

by Wentao Liang

In _hl_mmu_v2_hr_map(), If hl_mmu_get_hop_pte_phys_addr() fail to get physical address, the return address will be set as U64_MAX. Hence, the return value of hl_mmu_get_hop_pte_phys_addr() must be checked to prevent invalid address access. Add error handling and propagate return code to caller function to fix this issue. Fixes: 8aa1e1e60553 ("habanalabs: add gaudi2 MMU support") Cc: stable(a)vger.kernel.org # v6.0+ Signed-off-by: Wentao Liang <vulab(a)iscas.ac.cn> --- drivers/accel/habanalabs/common/mmu/mmu_v2_hr.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/accel/habanalabs/common/mmu/mmu_v2_hr.c b/drivers/accel/habanalabs/common/mmu/mmu_v2_hr.c index 31507b2a431b..cdade07e22c5 100644 --- a/drivers/accel/habanalabs/common/mmu/mmu_v2_hr.c +++ b/drivers/accel/habanalabs/common/mmu/mmu_v2_hr.c @@ -253,6 +253,11 @@ static int _hl_mmu_v2_hr_map(struct hl_ctx *ctx, hop_pte_phys_addr[i] = hl_mmu_get_hop_pte_phys_addr(ctx, mmu_prop, i, hops_pgt_info[i]->phys_addr, scrambled_virt_addr); + if (hop_pte_phys_addr[i] == U64_MAX) { + rc = -EINVAL; + goto err; + } + curr_pte = *(u64 *) (uintptr_t) hl_mmu_hr_pte_phys_to_virt(ctx, hops_pgt_info[i], hop_pte_phys_addr[i], ctx->hdev->asic_prop.pmmu.hop_table_size); -- 2.42.0.windows.2

9 months, 1 week

[PATCH 6.12] drm/amd/display: Don't write DP_MSTM_CTRL after LT

by Thadeu Lima de Souza Cascardo

From: Wayne Lin <Wayne.Lin(a)amd.com> [ Upstream commit bc068194f548ef1f230d96c4398046bf59165992 ] [Why] Observe after suspend/resme, we can't light up mst monitors under specific mst hub. The reason is that driver still writes DPCD DP_MSTM_CTRL after LT. It's forbidden even we write the same value for that dpcd register. [How] We already resume the mst branch device dpcd settings during resume_mst_branch_status(). Leverage drm_dp_mst_topology_queue_probe() to only probe the topology, not calling drm_dp_mst_topology_mgr_resume() which will set DP_MSTM_CTRL as well. Reviewed-by: Jerry Zuo <jerry.zuo(a)amd.com> Signed-off-by: Wayne Lin <Wayne.Lin(a)amd.com> Signed-off-by: Zaeem Mohamed <zaeem.mohamed(a)amd.com> Tested-by: Daniel Wheeler <daniel.wheeler(a)amd.com> Signed-off-by: Alex Deucher <alexander.deucher(a)amd.com> Signed-off-by: Thadeu Lima de Souza Cascardo <cascardo(a)igalia.com> --- .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 16 ++++------------ 1 file changed, 4 insertions(+), 12 deletions(-) diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c index d9a3917d207e..c4c6538eabae 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -3231,8 +3231,7 @@ static int dm_resume(void *handle) struct dm_atomic_state *dm_state = to_dm_atomic_state(dm->atomic_obj.state); enum dc_connection_type new_connection_type = dc_connection_none; struct dc_state *dc_state; - int i, r, j, ret; - bool need_hotplug = false; + int i, r, j; struct dc_commit_streams_params commit_params = {}; if (dm->dc->caps.ips_support) { @@ -3427,23 +3426,16 @@ static int dm_resume(void *handle) aconnector->mst_root) continue; - ret = drm_dp_mst_topology_mgr_resume(&aconnector->mst_mgr, true); - - if (ret < 0) { - dm_helpers_dp_mst_stop_top_mgr(aconnector->dc_link->ctx, - aconnector->dc_link); - need_hotplug = true; - } + drm_dp_mst_topology_queue_probe(&aconnector->mst_mgr); } drm_connector_list_iter_end(&iter); - if (need_hotplug) - drm_kms_helper_hotplug_event(ddev); - amdgpu_dm_irq_resume_late(adev); amdgpu_dm_smu_write_watermarks_table(adev); + drm_kms_helper_hotplug_event(ddev); + return 0; } -- 2.47.2

9 months, 1 week

Request to cherry-pick a few ARM mm fixes

by Munehisa Kamata

Hello stable maintainers, The following mainline commits fix a certain stack unwinding issue on ARM, which had existed till v6.9. Could you please cherry-pick them for 5.4.y - 6.6.y? 169f9102f919 ("ARM: 9350/1: fault: Implement copy_from_kernel_nofault_allowed()") 8f09b8b4fa58 ("ARM: 9351/1: fault: Add "cut here" line for prefetch aborts") 3ccea4784fdd ("ARM: Remove address checking for MMUless devices") Confirmed these cleanly apply and build in all the branches. Thanks, Munehisa

9 months, 1 week

Upstream patch pairing request

by Terry Junge

Hi Greg and Sasha, The following two patches have now been applied to mainline. commit 486f6205c233da1baa309bde5f634eb1f8319a33 ALSA: usb-audio: Add quirk for Plantronics headsets to fix control names commit 9821709af892be9fbf4ee9a50b2f3e0604295ce0 HID: hid-plantronics: Add mic mute mapping and generalize quirks It would be ideal if both of these patches could flow upstream together although neither actually depends on the other. A verbose description of the user experience with neither, one, the other, or both patches applied can be found here: https://lore.kernel.org/all/f107e133-c536-43e5-bd4f-4fcb8a4b4c7f@cosmicgizm… Let me know if there is anything I can do to help. Thanks, Terry

9 months, 1 week

[PATCH] batman-adv: batman-adv: handle tvlv unicast send errors

by Wentao Liang

In batadv_tvlv_unicast_send(), the return value of batadv_send_skb_to_orig() is ignored. This could silently drop send failures, making it difficult to detect connectivity issues. Add error checking for batadv_send_skb_to_orig() and log failures via batadv_dbg() to improve error visibility. Fixes: 1ad5bcb2a032 ("batman-adv: Consume skb in batadv_send_skb_to_orig") Cc: stable(a)vger.kernel.org # 4.10+ Signed-off-by: Wentao Liang <vulab(a)iscas.ac.cn> --- net/batman-adv/tvlv.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/net/batman-adv/tvlv.c b/net/batman-adv/tvlv.c index 2a583215d439..f081136cc5b7 100644 --- a/net/batman-adv/tvlv.c +++ b/net/batman-adv/tvlv.c @@ -625,6 +625,7 @@ void batadv_tvlv_unicast_send(struct batadv_priv *bat_priv, const u8 *src, unsigned char *tvlv_buff; unsigned int tvlv_len; ssize_t hdr_len = sizeof(*unicast_tvlv_packet); + int r; orig_node = batadv_orig_hash_find(bat_priv, dst); if (!orig_node) @@ -657,7 +658,10 @@ void batadv_tvlv_unicast_send(struct batadv_priv *bat_priv, const u8 *src, tvlv_buff += sizeof(*tvlv_hdr); memcpy(tvlv_buff, tvlv_value, tvlv_value_len); - batadv_send_skb_to_orig(skb, orig_node, NULL); + r = batadv_send_skb_to_orig(skb, orig_node, NULL); + if (r != NET_XMIT_SUCCESS) + batadv_dbg(BATADV_DBG_TP_METER, bat_priv, + "Fail to send the ack."); out: batadv_orig_node_put(orig_node); } -- 2.42.0.windows.2

9 months, 1 week

[PATCH] drm/i915/gvt: fix unterminated-string-initialization warning

by Jani Nikula

Initializing const char opregion_signature[16] = OPREGION_SIGNATURE (which is "IntelGraphicsMem") drops the NUL termination of the string. This is intentional, but the compiler doesn't know this. Switch to initializing header->signature directly from the string litaral, with sizeof destination rather than source. We don't treat the signature as a string other than for initialization; it's really just a blob of binary data. Add a static assert for good measure to cross-check the sizes. Reported-by: Kees Cook <kees(a)kernel.org> Closes: https://lore.kernel.org/r/20250310222355.work.417-kees@kernel.org Closes: https://gitlab.freedesktop.org/drm/i915/kernel/-/issues/13934 Tested-by: Nicolas Chauvet <kwizart(a)gmail.com> Tested-by: Damian Tometzki <damian(a)riscv-rocks.de> Cc: stable(a)vger.kernel.org Signed-off-by: Jani Nikula <jani.nikula(a)intel.com> --- drivers/gpu/drm/i915/gvt/opregion.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/i915/gvt/opregion.c b/drivers/gpu/drm/i915/gvt/opregion.c index 509f9ccae3a9..dbad4d853d3a 100644 --- a/drivers/gpu/drm/i915/gvt/opregion.c +++ b/drivers/gpu/drm/i915/gvt/opregion.c @@ -222,7 +222,6 @@ int intel_vgpu_init_opregion(struct intel_vgpu *vgpu) u8 *buf; struct opregion_header *header; struct vbt v; - const char opregion_signature[16] = OPREGION_SIGNATURE; gvt_dbg_core("init vgpu%d opregion\n", vgpu->id); vgpu_opregion(vgpu)->va = (void *)__get_free_pages(GFP_KERNEL | @@ -236,8 +235,10 @@ int intel_vgpu_init_opregion(struct intel_vgpu *vgpu) /* emulated opregion with VBT mailbox only */ buf = (u8 *)vgpu_opregion(vgpu)->va; header = (struct opregion_header *)buf; - memcpy(header->signature, opregion_signature, - sizeof(opregion_signature)); + + static_assert(sizeof(header->signature) == sizeof(OPREGION_SIGNATURE) - 1); + memcpy(header->signature, OPREGION_SIGNATURE, sizeof(header->signature)); + header->size = 0x8; header->opregion_ver = 0x02000000; header->mboxes = MBOX_VBT; -- 2.39.5

9 months, 1 week

[PATCH] drm/imagination: fix firmware memory leaks

by Brendan King via B4 Relay

From: Brendan King <Brendan.King(a)imgtec.com> Free the memory used to hold the results of firmware image processing when the module is unloaded. Fix the related issue of the same memory being leaked if processing of the firmware image fails during module load. Ensure all firmware GEM objects are destroyed if firmware image processing fails. Fixes memory leaks on powervr module unload detected by Kmemleak: unreferenced object 0xffff000042e20000 (size 94208): comm "modprobe", pid 470, jiffies 4295277154 hex dump (first 32 bytes): 02 ae 7f ed bf 45 84 00 3c 5b 1f ed 9f 45 45 05 .....E..<[...EE. d5 4f 5d 14 6c 00 3d 23 30 d0 3a 4a 66 0e 48 c8 .O].l.=#0.:Jf.H. backtrace (crc dd329dec): kmemleak_alloc+0x30/0x40 ___kmalloc_large_node+0x140/0x188 __kmalloc_large_node_noprof+0x2c/0x13c __kmalloc_noprof+0x48/0x4c0 pvr_fw_init+0xaa4/0x1f50 [powervr] unreferenced object 0xffff000042d20000 (size 20480): comm "modprobe", pid 470, jiffies 4295277154 hex dump (first 32 bytes): 00 00 00 00 00 00 00 00 09 00 00 00 0b 00 00 00 ................ 00 00 00 00 00 00 00 00 07 00 00 00 08 00 00 00 ................ backtrace (crc 395b02e3): kmemleak_alloc+0x30/0x40 ___kmalloc_large_node+0x140/0x188 __kmalloc_large_node_noprof+0x2c/0x13c __kmalloc_noprof+0x48/0x4c0 pvr_fw_init+0xb0c/0x1f50 [powervr] Cc: stable(a)vger.kernel.org Fixes: cc1aeedb98ad ("drm/imagination: Implement firmware infrastructure and META FW support") Signed-off-by: Brendan King <brendan.king(a)imgtec.com> --- drivers/gpu/drm/imagination/pvr_fw.c | 27 ++++++++++++++++++++------- 1 file changed, 20 insertions(+), 7 deletions(-) diff --git a/drivers/gpu/drm/imagination/pvr_fw.c b/drivers/gpu/drm/imagination/pvr_fw.c index 3debc9870a82ae7de9b2dc173df84c466c137bb3..d09c4c68411627714c14dee5ed4e61b07baca1ba 100644 --- a/drivers/gpu/drm/imagination/pvr_fw.c +++ b/drivers/gpu/drm/imagination/pvr_fw.c @@ -732,7 +732,7 @@ pvr_fw_process(struct pvr_device *pvr_dev) fw_mem->core_data, fw_mem->core_code_alloc_size); if (err) - goto err_free_fw_core_data_obj; + goto err_free_kdata; memcpy(fw_code_ptr, fw_mem->code, fw_mem->code_alloc_size); memcpy(fw_data_ptr, fw_mem->data, fw_mem->data_alloc_size); @@ -742,10 +742,14 @@ pvr_fw_process(struct pvr_device *pvr_dev) memcpy(fw_core_data_ptr, fw_mem->core_data, fw_mem->core_data_alloc_size); /* We're finished with the firmware section memory on the CPU, unmap. */ - if (fw_core_data_ptr) + if (fw_core_data_ptr) { pvr_fw_object_vunmap(fw_mem->core_data_obj); - if (fw_core_code_ptr) + fw_core_data_ptr = NULL; + } + if (fw_core_code_ptr) { pvr_fw_object_vunmap(fw_mem->core_code_obj); + fw_core_code_ptr = NULL; + } pvr_fw_object_vunmap(fw_mem->data_obj); fw_data_ptr = NULL; pvr_fw_object_vunmap(fw_mem->code_obj); @@ -753,7 +757,7 @@ pvr_fw_process(struct pvr_device *pvr_dev) err = pvr_fw_create_fwif_connection_ctl(pvr_dev); if (err) - goto err_free_fw_core_data_obj; + goto err_free_kdata; return 0; @@ -763,13 +767,16 @@ pvr_fw_process(struct pvr_device *pvr_dev) kfree(fw_mem->data); kfree(fw_mem->code); -err_free_fw_core_data_obj: if (fw_core_data_ptr) - pvr_fw_object_unmap_and_destroy(fw_mem->core_data_obj); + pvr_fw_object_vunmap(fw_mem->core_data_obj); + if (fw_mem->core_data_obj) + pvr_fw_object_destroy(fw_mem->core_data_obj); err_free_fw_core_code_obj: if (fw_core_code_ptr) - pvr_fw_object_unmap_and_destroy(fw_mem->core_code_obj); + pvr_fw_object_vunmap(fw_mem->core_code_obj); + if (fw_mem->core_code_obj) + pvr_fw_object_destroy(fw_mem->core_code_obj); err_free_fw_data_obj: if (fw_data_ptr) @@ -836,6 +843,12 @@ pvr_fw_cleanup(struct pvr_device *pvr_dev) struct pvr_fw_mem *fw_mem = &pvr_dev->fw_dev.mem; pvr_fw_fini_fwif_connection_ctl(pvr_dev); + + kfree(fw_mem->core_data); + kfree(fw_mem->core_code); + kfree(fw_mem->data); + kfree(fw_mem->code); + if (fw_mem->core_code_obj) pvr_fw_object_destroy(fw_mem->core_code_obj); if (fw_mem->core_data_obj) --- base-commit: 96c85e428ebaeacd2c640eba075479ab92072ccd change-id: 20250318-ddkopsrc-1339-firmware-related-memory-leak-on-module-unload-c18a9a4fd0db Best regards, -- Brendan King <Brendan.King(a)imgtec.com>

9 months, 1 week

[PATCH] drm/imagination: take paired job reference

by Brendan King via B4 Relay

From: Brendan King <Brendan.King(a)imgtec.com> For paired jobs, have the fragment job take a reference on the geometry job, so that the geometry job cannot be freed until the fragment job has finished with it. The geometry job structure is accessed when the fragment job is being prepared by the GPU scheduler. Taking the reference prevents the geometry job being freed until the fragment job no longer requires it. Fixes a use after free bug detected by KASAN: [ 124.256386] BUG: KASAN: slab-use-after-free in pvr_queue_prepare_job+0x108/0x868 [powervr] [ 124.264893] Read of size 1 at addr ffff0000084cb960 by task kworker/u16:4/63 Cc: stable(a)vger.kernel.org Fixes: eaf01ee5ba28 ("drm/imagination: Implement job submission and scheduling") Signed-off-by: Brendan King <brendan.king(a)imgtec.com> --- drivers/gpu/drm/imagination/pvr_job.c | 7 +++++++ drivers/gpu/drm/imagination/pvr_queue.c | 4 ++++ 2 files changed, 11 insertions(+) diff --git a/drivers/gpu/drm/imagination/pvr_job.c b/drivers/gpu/drm/imagination/pvr_job.c index 1cdb3cfd058d7db573337a2b4f6895ee4922f9a9..59b334d094fa826f26668d98561e956ec9c51428 100644 --- a/drivers/gpu/drm/imagination/pvr_job.c +++ b/drivers/gpu/drm/imagination/pvr_job.c @@ -671,6 +671,13 @@ pvr_jobs_link_geom_frag(struct pvr_job_data *job_data, u32 *job_count) geom_job->paired_job = frag_job; frag_job->paired_job = geom_job; + /* The geometry job pvr_job structure is used when the fragment + * job is being prepared by the GPU scheduler. Have the fragment + * job hold a reference on the geometry job to prevent it being + * freed until the fragment job has finished with it. + */ + pvr_job_get(geom_job); + /* Skip the fragment job we just paired to the geometry job. */ i++; } diff --git a/drivers/gpu/drm/imagination/pvr_queue.c b/drivers/gpu/drm/imagination/pvr_queue.c index 21c185d18bb2e0569bd6e12832a74e38137bd48a..6431f6b654a2e60b86a46bd8571eb9f8133c4b53 100644 --- a/drivers/gpu/drm/imagination/pvr_queue.c +++ b/drivers/gpu/drm/imagination/pvr_queue.c @@ -856,6 +856,10 @@ static void pvr_queue_free_job(struct drm_sched_job *sched_job) struct pvr_job *job = container_of(sched_job, struct pvr_job, base); drm_sched_job_cleanup(sched_job); + + if (job->type == DRM_PVR_JOB_TYPE_FRAGMENT && job->paired_job) + pvr_job_put(job->paired_job); + job->paired_job = NULL; pvr_job_put(job); } --- base-commit: 96c85e428ebaeacd2c640eba075479ab92072ccd change-id: 20250318-ddkopsrc-1337-use-after-free-in-pvr_queue_prepare_job-16fbc74b0c20 Best regards, -- Brendan King <Brendan.King(a)imgtec.com>

9 months, 1 week

Jump to page:

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror April 2025