July 2025 - Linux-stable-mirror

[PATCH v3 1/9] vhost/vsock: Avoid allocating arbitrarily-sized SKBs

by Will Deacon

vhost_vsock_alloc_skb() returns NULL for packets advertising a length larger than VIRTIO_VSOCK_MAX_PKT_BUF_SIZE in the packet header. However, this is only checked once the SKB has been allocated and, if the length in the packet header is zero, the SKB may not be freed immediately. Hoist the size check before the SKB allocation so that an iovec larger than VIRTIO_VSOCK_MAX_PKT_BUF_SIZE + the header size is rejected outright. The subsequent check on the length field in the header can then simply check that the allocated SKB is indeed large enough to hold the packet. Cc: <stable(a)vger.kernel.org> Fixes: 71dc9ec9ac7d ("virtio/vsock: replace virtio_vsock_pkt with sk_buff") Signed-off-by: Will Deacon <will(a)kernel.org> --- drivers/vhost/vsock.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/vhost/vsock.c b/drivers/vhost/vsock.c index 802153e23073..66a0f060770e 100644 --- a/drivers/vhost/vsock.c +++ b/drivers/vhost/vsock.c @@ -344,6 +344,9 @@ vhost_vsock_alloc_skb(struct vhost_virtqueue *vq, len = iov_length(vq->iov, out); + if (len > VIRTIO_VSOCK_MAX_PKT_BUF_SIZE + VIRTIO_VSOCK_SKB_HEADROOM) + return NULL; + /* len contains both payload and hdr */ skb = virtio_vsock_alloc_skb(len, GFP_KERNEL); if (!skb) @@ -367,8 +370,7 @@ vhost_vsock_alloc_skb(struct vhost_virtqueue *vq, return skb; /* The pkt is too big or the length in the header is invalid */ - if (payload_len > VIRTIO_VSOCK_MAX_PKT_BUF_SIZE || - payload_len + sizeof(*hdr) > len) { + if (payload_len + sizeof(*hdr) > len) { kfree_skb(skb); return NULL; } -- 2.50.0.727.gbf7dc18ff4-goog

4 months, 1 week

2
1
0 0

[PATCH net 0/2] selftests: mptcp: connect: cover alt modes

by Matthieu Baerts (NGI0)

mptcp_connect.sh can be executed manually with "-m <MODE>" and "-C" to make sure everything works as expected when using "mmap" and "sendfile" modes instead of "poll", and with the MPTCP checksum support. These modes should be validated, but they are not when the selftests are executed via the kselftest helpers. It means that most CIs validating these selftests, like NIPA for the net development trees and LKFT for the stable ones, are not covering these modes. To fix that, new test programs have been added, simply calling mptcp_connect.sh with the right parameters. The first patch can be backported up to v5.6, and the second one up to v5.14. Signed-off-by: Matthieu Baerts (NGI0) <matttbe(a)kernel.org> --- Matthieu Baerts (NGI0) (2): selftests: mptcp: connect: also cover alt modes selftests: mptcp: connect: also cover checksum tools/testing/selftests/net/mptcp/Makefile | 3 ++- tools/testing/selftests/net/mptcp/mptcp_connect_checksum.sh | 4 ++++ tools/testing/selftests/net/mptcp/mptcp_connect_mmap.sh | 4 ++++ tools/testing/selftests/net/mptcp/mptcp_connect_sendfile.sh | 4 ++++ 4 files changed, 14 insertions(+), 1 deletion(-) --- base-commit: b640daa2822a39ff76e70200cb2b7b892b896dce change-id: 20250714-net-mptcp-sft-connect-alt-c1aaf073ef4e Best regards, -- Matthieu Baerts (NGI0) <matttbe(a)kernel.org>

4 months, 1 week

2
3
0 0

[PATCH 5.4] net: ipv6: Discard next-hop MTU less than minimum link MTU

by WangYuli

From: Georg Kohmann <geokohma(a)cisco.com> [ Upstream commit 4a65dff81a04f874fa6915c7f069b4dc2c4010e4 ] When a ICMPV6_PKT_TOOBIG report a next-hop MTU that is less than the IPv6 minimum link MTU, the estimated path MTU is reduced to the minimum link MTU. This behaviour breaks TAHI IPv6 Core Conformance Test v6LC4.1.6: Packet Too Big Less than IPv6 MTU. Referring to RFC 8201 section 4: "If a node receives a Packet Too Big message reporting a next-hop MTU that is less than the IPv6 minimum link MTU, it must discard it. A node must not reduce its estimate of the Path MTU below the IPv6 minimum link MTU on receipt of a Packet Too Big message." Drop the path MTU update if reported MTU is less than the minimum link MTU. Signed-off-by: Georg Kohmann <geokohma(a)cisco.com> Signed-off-by: Jakub Kicinski <kuba(a)kernel.org> Signed-off-by: WangYuli <wangyuli(a)uniontech.com> --- net/ipv6/route.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/net/ipv6/route.c b/net/ipv6/route.c index b321cec71127..93bae134e730 100644 --- a/net/ipv6/route.c +++ b/net/ipv6/route.c @@ -2765,7 +2765,8 @@ static void __ip6_rt_update_pmtu(struct dst_entry *dst, const struct sock *sk, if (confirm_neigh) dst_confirm_neigh(dst, daddr); - mtu = max_t(u32, mtu, IPV6_MIN_MTU); + if (mtu < IPV6_MIN_MTU) + return; if (mtu >= dst_mtu(dst)) return; -- 2.50.0

4 months, 1 week

1
0
0 0

[PATCH 5.4/5.10/5.15/6.1/6.6] Input: atkbd - do not skip atkbd_deactivate() when skipping ATKBD_CMD_GETID

by Wang Hai

From: Hans de Goede <hdegoede(a)redhat.com> commit 9cf6e24c9fbf17e52de9fff07f12be7565ea6d61 upstream. After commit 936e4d49ecbc ("Input: atkbd - skip ATKBD_CMD_GETID in translated mode") not only the getid command is skipped, but also the de-activating of the keyboard at the end of atkbd_probe(), potentially re-introducing the problem fixed by commit be2d7e4233a4 ("Input: atkbd - fix multi-byte scancode handling on reconnect"). Make sure multi-byte scancode handling on reconnect is still handled correctly by not skipping the atkbd_deactivate() call. Fixes: 936e4d49ecbc ("Input: atkbd - skip ATKBD_CMD_GETID in translated mode") Tested-by: Paul Menzel <pmenzel(a)molgen.mpg.de> Signed-off-by: Hans de Goede <hdegoede(a)redhat.com> Link: https://lore.kernel.org/r/20240126160724.13278-3-hdegoede@redhat.com Signed-off-by: Dmitry Torokhov <dmitry.torokhov(a)gmail.com> Signed-off-by: Wang Hai <wanghai38(a)huawei.com> --- drivers/input/keyboard/atkbd.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/input/keyboard/atkbd.c b/drivers/input/keyboard/atkbd.c index b3a856333d4e..de59fc1a24bc 100644 --- a/drivers/input/keyboard/atkbd.c +++ b/drivers/input/keyboard/atkbd.c @@ -805,11 +805,11 @@ static int atkbd_probe(struct atkbd *atkbd) "keyboard reset failed on %s\n", ps2dev->serio->phys); if (atkbd_skip_getid(atkbd)) { atkbd->id = 0xab83; - return 0; + goto deactivate_kbd; } /* * Then we check the keyboard ID. We should get 0xab83 under normal conditions. * Some keyboards report different values, but the first byte is always 0xab or @@ -842,10 +842,11 @@ static int atkbd_probe(struct atkbd *atkbd) "NCD terminal keyboards are only supported on non-translating controllers. " "Use i8042.direct=1 to disable translation.\n"); return -1; } +deactivate_kbd: /* * Make sure nothing is coming from the keyboard and disturbs our * internal state. */ if (!atkbd_skip_deactivate) -- 2.17.1

4 months, 1 week

2
1
0 0

[PATCH] media: pci: intel: Balance device refcount when destroying devices

by Ma Ke

Using ipu_bridge_get_ivsc_csi_dev() to locate the device could cause an imbalance in the device's reference count. ipu_bridge_get_ivsc_csi_dev() calls device_find_child_by_name() to implement the localization, and device_find_child_by_name() calls an implicit get_device() to increment the device's reference count before returning the pointer. Throughout the entire implementation process, no mechanism releases resources properly. This leads to a memory leak because the reference count of the device is never decremented. As the comment of device_find_child_by_name() says, 'NOTE: you will need to drop the reference with put_device() after use'. Found by code review. Cc: stable(a)vger.kernel.org Fixes: c66821f381ae ("media: pci: intel: Add IVSC support for IPU bridge driver") Signed-off-by: Ma Ke <make24(a)iscas.ac.cn> --- drivers/media/pci/intel/ipu-bridge.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/media/pci/intel/ipu-bridge.c b/drivers/media/pci/intel/ipu-bridge.c index 83e682e1a4b7..f8b4672accab 100644 --- a/drivers/media/pci/intel/ipu-bridge.c +++ b/drivers/media/pci/intel/ipu-bridge.c @@ -192,6 +192,7 @@ static int ipu_bridge_check_ivsc_dev(struct ipu_sensor *sensor, sensor->csi_dev = csi_dev; sensor->ivsc_adev = adev; + put_device(csi_dev); } return 0; -- 2.25.1

4 months, 1 week

2
1
0 0

[PATCH] KVM: SEV: Enforce minimum GHCB version requirement for SEV-SNP guests

by Nikunj A Dadhania

Require a minimum GHCB version of 2 when starting SEV-SNP guests through KVM_SEV_INIT2. When a VMM attempts to start an SEV-SNP guest with an incompatible GHCB version (less than 2), reject the request early rather than allowing the guest to start with an incorrect protocol version and fail later. Fixes: 4af663c2f64a ("KVM: SEV: Allow per-guest configuration of GHCB protocol version") Cc: Thomas Lendacky <thomas.lendacky(a)amd.com> Cc: Sean Christopherson <seanjc(a)google.com> Cc: Michael Roth <michael.roth(a)amd.com> Cc: stable(a)vger.kernel.org Signed-off-by: Nikunj A Dadhania <nikunj(a)amd.com> --- arch/x86/kvm/svm/sev.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/arch/x86/kvm/svm/sev.c b/arch/x86/kvm/svm/sev.c index a12e78b67466..91d06fb91ba2 100644 --- a/arch/x86/kvm/svm/sev.c +++ b/arch/x86/kvm/svm/sev.c @@ -435,6 +435,9 @@ static int __sev_guest_init(struct kvm *kvm, struct kvm_sev_cmd *argp, if (unlikely(sev->active)) return -EINVAL; + if (snp_active && data->ghcb_version && data->ghcb_version < 2) + return -EINVAL; + sev->active = true; sev->es_active = es_active; sev->vmsa_features = data->vmsa_features; -- 2.43.0

4 months, 1 week

3
6
0 0

[REGRESSION] thunderbolt: Fix a logic error in wake on connect

by Alyssa Ross

On Fri, Apr 11, 2025 at 10:14:44AM -0500, Mario Limonciello wrote: > From: Mario Limonciello <mario.limonciello(a)amd.com> > > commit a5cfc9d65879c ("thunderbolt: Add wake on connect/disconnect > on USB4 ports") introduced a sysfs file to control wake up policy > for a given USB4 port that defaulted to disabled. > > However when testing commit 4bfeea6ec1c02 ("thunderbolt: Use wake > on connect and disconnect over suspend") I found that it was working > even without making changes to the power/wakeup file (which defaults > to disabled). This is because of a logic error doing a bitwise or > of the wake-on-connect flag with device_may_wakeup() which should > have been a logical AND. > > Adjust the logic so that policy is only applied when wakeup is > actually enabled. > > Fixes: a5cfc9d65879c ("thunderbolt: Add wake on connect/disconnect on USB4 ports") > Signed-off-by: Mario Limonciello <mario.limonciello(a)amd.com> Hi! There have been a couple of reports of a Thunderbolt regression in recent stable kernels, and one reporter has now bisected it to this change: • https://bugzilla.kernel.org/show_bug.cgi?id=220284 • https://github.com/NixOS/nixpkgs/issues/420730 Both reporters are CCed, and say it starts working after the module is reloaded. Link: https://lore.kernel.org/r/bug-220284-208809@https.bugzilla.kernel.org%2F/ (for regzbot) > --- > drivers/thunderbolt/usb4.c | 4 ++-- > 1 file changed, 2 insertions(+), 2 deletions(-) > > diff --git a/drivers/thunderbolt/usb4.c b/drivers/thunderbolt/usb4.c > index e51d01671d8e7..3e96f1afd4268 100644 > --- a/drivers/thunderbolt/usb4.c > +++ b/drivers/thunderbolt/usb4.c > @@ -440,10 +440,10 @@ int usb4_switch_set_wake(struct tb_switch *sw, unsigned int flags) > bool configured = val & PORT_CS_19_PC; > usb4 = port->usb4; > > - if (((flags & TB_WAKE_ON_CONNECT) | > + if (((flags & TB_WAKE_ON_CONNECT) && > device_may_wakeup(&usb4->dev)) && !configured) > val |= PORT_CS_19_WOC; > - if (((flags & TB_WAKE_ON_DISCONNECT) | > + if (((flags & TB_WAKE_ON_DISCONNECT) && > device_may_wakeup(&usb4->dev)) && configured) > val |= PORT_CS_19_WOD; > if ((flags & TB_WAKE_ON_USB4) && configured) > -- > 2.43.0 >

4 months, 1 week

3
5
0 0

Re: [PATCH 6.6 087150] Input atkbd - skip ATKBD_CMD_GETID in translated mode

by Wang Hai

On 1970/1/1 8:00, wrote: > 6.6-stable review patch. If anyone has any objections, please let me know. > > ------------------ > > From Hans de Goede hdegoede(a)redhat.com > > [ Upstream commit 936e4d49ecbc8c404790504386e1422b599dec39 ] > > There have been multiple reports of keyboard issues on recent laptop models > which can be worked around by setting i8042.dumbkbd, with the downside > being this breaks the capslock LED. > > It seems that these issues are caused by recent laptops getting confused by > ATKBD_CMD_GETID. Rather then adding and endless growing list of quirks for > this, just skip ATKBD_CMD_GETID alltogether on laptops in translated mode. > > The main goal of sending ATKBD_CMD_GETID is to skip binding to ps2 > micetouchpads and those are never used in translated mode. > > Examples of laptop models which benefit from skipping ATKBD_CMD_GETID > > HP Laptop 15s-fq2xxx, HP laptop 15s-fq4xxx and HP Laptop 15-dy2xxx > models the kbd stops working for the first 2 - 5 minutes after boot > (waiting for EC watchdog reset) > > On HP Spectre x360 13-aw2xxx atkbd fails to probe the keyboard > > At least 9 different Lenovo models have issues with ATKBD_CMD_GETID, see > httpsgithub.comyescallopatkbd-nogetid > > This has been tested on > > 1. A MSI B550M PRO-VDH WIFI desktop, where the i8042 controller is not > in translated mode when no keyboard is plugged in and with a ps2 kbd > a AT Translated Set 2 keyboard devinputevent# node shows up > > 2. A Lenovo ThinkPad X1 Yoga gen 8 (always has a translated set 2 keyboard) > > Reported-by Shang Ye yesh25(a)mail2.sysu.edu.cn > Closes httpslore.kernel.orglinux-input886D6167733841AE+20231017135318.11142-1-yesh25(a)mail2.sysu.edu.cn > Closes httpsgithub.comyescallopatkbd-nogetid > Reported-by gurevitch mail(a)gurevit.ch > Closes httpslore.kernel.orglinux-input2iAJTwqZV6lQs26cTb38RNYqxvsink6SRmrZ5h0cBUSuf9NT0tZTsf9fEAbbto2maavHJEOP8GA1evlKa6xjKOsaskDhtJWxjcnrgPigzVo=(a)gurevit.ch > Reported-by Egor Ignatov egori(a)altlinux.org > Closes httpslore.kernel.orgall20210609073333.8425-1-egori(a)altlinux.org > Reported-by Anton Zhilyaev anton(a)cpp.in > Closes httpslore.kernel.orglinux-input20210201160336.16008-1-anton(a)cpp.in > Closes httpsbugzilla.redhat.comshow_bug.cgiid=2086156 > Signed-off-by Hans de Goede hdegoede(a)redhat.com > Link httpslore.kernel.orgr20231115174625.7462-1-hdegoede(a)redhat.com > Signed-off-by Dmitry Torokhov dmitry.torokhov(a)gmail.com > Signed-off-by Sasha Levin sashal(a)kernel.org > --- Hi, Hans I noticed there's a subsequent bugfix [1] for this patch, but it hasn't been merged into the stable-6.6 branch. Based on the bugfix description, the issue should exist there as well. Would you like this patch to be merged into the stable-6.6 branch?" [1] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?…

4 months, 1 week

2
2
0 0

[PATCH v6.1] vhost-scsi: protect vq->log_used with vq->mutex

by Xinyu Zheng

From: Dongli Zhang <dongli.zhang(a)oracle.com> [ Upstream commit f591cf9fce724e5075cc67488c43c6e39e8cbe27 ] The vhost-scsi completion path may access vq->log_base when vq->log_used is already set to false. vhost-thread QEMU-thread vhost_scsi_complete_cmd_work() -> vhost_add_used() -> vhost_add_used_n() if (unlikely(vq->log_used)) QEMU disables vq->log_used via VHOST_SET_VRING_ADDR. mutex_lock(&vq->mutex); vq->log_used = false now! mutex_unlock(&vq->mutex); QEMU gfree(vq->log_base) log_used() -> log_write(vq->log_base) Assuming the VMM is QEMU. The vq->log_base is from QEMU userpace and can be reclaimed via gfree(). As a result, this causes invalid memory writes to QEMU userspace. The control queue path has the same issue. Cc: stable(a)vger.kernel.org#6.1.x Cc: gregkh(a)linuxfoundation.org Signed-off-by: Dongli Zhang <dongli.zhang(a)oracle.com> Acked-by: Jason Wang <jasowang(a)redhat.com> Reviewed-by: Mike Christie <michael.christie(a)oracle.com> Message-Id: <20250403063028.16045-2-dongli.zhang(a)oracle.com> Signed-off-by: Michael S. Tsirkin <mst(a)redhat.com> Signed-off-by: Sasha Levin <sashal(a)kernel.org> [ Conflicts in drivers/vhost/scsi.c bacause vhost_scsi_complete_cmd_work() has been refactored. ] Signed-off-by: Xinyu Zheng <zhengxinyu6(a)huawei.com> --- drivers/vhost/scsi.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/drivers/vhost/scsi.c b/drivers/vhost/scsi.c index 3077cb9d58d6..87f2f56fd20a 100644 --- a/drivers/vhost/scsi.c +++ b/drivers/vhost/scsi.c @@ -568,8 +568,10 @@ static void vhost_scsi_complete_cmd_work(struct vhost_work *work) ret = copy_to_iter(&v_rsp, sizeof(v_rsp), &iov_iter); if (likely(ret == sizeof(v_rsp))) { struct vhost_scsi_virtqueue *q; - vhost_add_used(cmd->tvc_vq, cmd->tvc_vq_desc, 0); q = container_of(cmd->tvc_vq, struct vhost_scsi_virtqueue, vq); + mutex_lock(&q->vq.mutex); + vhost_add_used(cmd->tvc_vq, cmd->tvc_vq_desc, 0); + mutex_unlock(&q->vq.mutex); vq = q - vs->vqs; __set_bit(vq, vs->compl_bitmap); } else @@ -1173,8 +1175,11 @@ static void vhost_scsi_tmf_resp_work(struct vhost_work *work) else resp_code = VIRTIO_SCSI_S_FUNCTION_REJECTED; + mutex_lock(&tmf->svq->vq.mutex); vhost_scsi_send_tmf_resp(tmf->vhost, &tmf->svq->vq, tmf->in_iovs, tmf->vq_desc, &tmf->resp_iov, resp_code); + mutex_unlock(&tmf->svq->vq.mutex); + vhost_scsi_release_tmf_res(tmf); } -- 2.34.1

4 months, 1 week

1
0
0 0

[PATCH v5.15] vhost-scsi: protect vq->log_used with vq->mutex

by Xinyu Zheng

From: Dongli Zhang <dongli.zhang(a)oracle.com> [ Upstream commit f591cf9fce724e5075cc67488c43c6e39e8cbe27 ] The vhost-scsi completion path may access vq->log_base when vq->log_used is already set to false. vhost-thread QEMU-thread vhost_scsi_complete_cmd_work() -> vhost_add_used() -> vhost_add_used_n() if (unlikely(vq->log_used)) QEMU disables vq->log_used via VHOST_SET_VRING_ADDR. mutex_lock(&vq->mutex); vq->log_used = false now! mutex_unlock(&vq->mutex); QEMU gfree(vq->log_base) log_used() -> log_write(vq->log_base) Assuming the VMM is QEMU. The vq->log_base is from QEMU userpace and can be reclaimed via gfree(). As a result, this causes invalid memory writes to QEMU userspace. The control queue path has the same issue. Cc: stable(a)vger.kernel.org#5.15.x Cc: gregkh(a)linuxfoundation.org Signed-off-by: Dongli Zhang <dongli.zhang(a)oracle.com> Acked-by: Jason Wang <jasowang(a)redhat.com> Reviewed-by: Mike Christie <michael.christie(a)oracle.com> Message-Id: <20250403063028.16045-2-dongli.zhang(a)oracle.com> Signed-off-by: Michael S. Tsirkin <mst(a)redhat.com> Signed-off-by: Sasha Levin <sashal(a)kernel.org> [ Conflicts in drivers/vhost/scsi.c bacause vhost_scsi_complete_cmd_work() has been refactored. ] Signed-off-by: Xinyu Zheng <zhengxinyu6(a)huawei.com> --- drivers/vhost/scsi.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/drivers/vhost/scsi.c b/drivers/vhost/scsi.c index f9930887fdd2..2c19fa02141d 100644 --- a/drivers/vhost/scsi.c +++ b/drivers/vhost/scsi.c @@ -563,8 +563,10 @@ static void vhost_scsi_complete_cmd_work(struct vhost_work *work) ret = copy_to_iter(&v_rsp, sizeof(v_rsp), &iov_iter); if (likely(ret == sizeof(v_rsp))) { struct vhost_scsi_virtqueue *q; - vhost_add_used(cmd->tvc_vq, cmd->tvc_vq_desc, 0); q = container_of(cmd->tvc_vq, struct vhost_scsi_virtqueue, vq); + mutex_lock(&q->vq.mutex); + vhost_add_used(cmd->tvc_vq, cmd->tvc_vq_desc, 0); + mutex_unlock(&q->vq.mutex); vq = q - vs->vqs; __set_bit(vq, signal); } else @@ -1166,8 +1168,11 @@ static void vhost_scsi_tmf_resp_work(struct vhost_work *work) else resp_code = VIRTIO_SCSI_S_FUNCTION_REJECTED; + mutex_lock(&tmf->svq->vq.mutex); vhost_scsi_send_tmf_resp(tmf->vhost, &tmf->svq->vq, tmf->in_iovs, tmf->vq_desc, &tmf->resp_iov, resp_code); + mutex_unlock(&tmf->svq->vq.mutex); + vhost_scsi_release_tmf_res(tmf); } -- 2.34.1

4 months, 1 week

1
0
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror July 2025