- Linux-stable-mirror - lists.linaro.org

[PATCH v4] KVM: x86: Add x2APIC "features" to control EOI broadcast suppression

by Khushit Shah

Add two flags for KVM_CAP_X2APIC_API to allow userspace to control support for Suppress EOI Broadcasts, which KVM completely mishandles. When x2APIC support was first added, KVM incorrectly advertised and "enabled" Suppress EOI Broadcast, without fully supporting the I/O APIC side of the equation, i.e. without adding directed EOI to KVM's in-kernel I/O APIC. That flaw was carried over to split IRQCHIP support, i.e. KVM advertised support for Suppress EOI Broadcasts irrespective of whether or not the userspace I/O APIC implementation supported directed EOIs. Even worse, KVM didn't actually suppress EOI broadcasts, i.e. userspace VMMs without support for directed EOI came to rely on the "spurious" broadcasts. KVM "fixed" the in-kernel I/O APIC implementation by completely disabling support for Suppress EOI Broadcasts in commit 0bcc3fb95b97 ("KVM: lapic: stop advertising DIRECTED_EOI when in-kernel IOAPIC is in use"), but didn't do anything to remedy userspace I/O APIC implementations. KVM's bogus handling of Suppress EOI Broadcast is problematic when the guest relies on interrupts being masked in the I/O APIC until well after the initial local APIC EOI. E.g. Windows with Credential Guard enabled handles interrupts in the following order: 1. Interrupt for L2 arrives. 2. L1 APIC EOIs the interrupt. 3. L1 resumes L2 and injects the interrupt. 4. L2 EOIs after servicing. 5. L1 performs the I/O APIC EOI. Because KVM EOIs the I/O APIC at step #2, the guest can get an interrupt storm, e.g. if the IRQ line is still asserted and userspace reacts to the EOI by re-injecting the IRQ, because the guest doesn't de-assert the line until step #4, and doesn't expect the interrupt to be re-enabled until step #5. Unfortunately, simply "fixing" the bug isn't an option, as KVM has no way of knowing if the userspace I/O APIC supports directed EOIs, i.e. suppressing EOI broadcasts would result in interrupts being stuck masked in the userspace I/O APIC due to step #5 being ignored by userspace. And fully disabling support for Suppress EOI Broadcast is also undesirable, as picking up the fix would require a guest reboot, *and* more importantly would change the virtual CPU model exposed to the guest without any buy-in from userspace. Add KVM_X2APIC_ENABLE_SUPPRESS_EOI_BROADCAST and KVM_X2APIC_DISABLE_SUPPRESS_EOI_BROADCAST flags to allow userspace to explicitly enable or disable support for Suppress EOI Broadcasts while using split IRQCHIP mode. This gives userspace control over the virtual CPU model exposed to the guest, as KVM should never have enabled support for Suppress EOI Broadcast without a userspace opt-in. Not setting either flag will result in legacy quirky behavior for backward compatibility. Note, Suppress EOI Broadcasts is defined only in Intel's SDM, not in AMD's APM. But the bit is writable on some AMD CPUs, e.g. Turin, and KVM's ABI is to support Directed EOI (KVM's name) irrespective of guest CPU vendor. Fixes: 7543a635aa09 ("KVM: x86: Add KVM exit for IOAPIC EOIs") Closes: https://lore.kernel.org/kvm/7D497EF1-607D-4D37-98E7-DAF95F099342@nutanix.com Cc: stable(a)vger.kernel.org Suggested-by: David Woodhouse <dwmw2(a)infradead.org> Co-developed-by: Sean Christopherson <seanjc(a)google.com> Signed-off-by: Sean Christopherson <seanjc(a)google.com> Signed-off-by: Khushit Shah <khushit.shah(a)nutanix.com> --- v4: - Add KVM_X2APIC_ENABLE_SUPPRESS_EOI_BROADCAST and KVM_X2APIC_DISABLE_SUPPRESS_EOI_BROADCAST flags to allow userspace to explicitly enable or disable support for Suppress EOI Broadcasts while using split IRQCHIP mode. After the inclusion of David Woodhouse's patch to support IOAPIC version 0x20, we can tweak the uAPI to support kernel IRQCHIP mode as well. Testing: - Setting both the flags fails with EINVAL. - Setting flags in kernel IRQCHIP mode fails with EINVAL. - Setting flags in split IRQCHIP mode succeeds and both the flags work as expected. --- Documentation/virt/kvm/api.rst | 27 ++++++++++++++++++++-- arch/x86/include/asm/kvm_host.h | 7 ++++++ arch/x86/include/uapi/asm/kvm.h | 6 +++-- arch/x86/kvm/lapic.c | 40 +++++++++++++++++++++++++++++++++ arch/x86/kvm/x86.c | 19 +++++++++++++--- 5 files changed, 92 insertions(+), 7 deletions(-) diff --git a/Documentation/virt/kvm/api.rst b/Documentation/virt/kvm/api.rst index 57061fa29e6a..b26528e0fec1 100644 --- a/Documentation/virt/kvm/api.rst +++ b/Documentation/virt/kvm/api.rst @@ -7800,8 +7800,10 @@ Will return -EBUSY if a VCPU has already been created. Valid feature flags in args[0] are:: - #define KVM_X2APIC_API_USE_32BIT_IDS (1ULL << 0) - #define KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK (1ULL << 1) + #define KVM_X2APIC_API_USE_32BIT_IDS (1ULL << 0) + #define KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK (1ULL << 1) + #define KVM_X2APIC_ENABLE_SUPPRESS_EOI_BROADCAST (1ULL << 2) + #define KVM_X2APIC_DISABLE_SUPPRESS_EOI_BROADCAST (1ULL << 3) Enabling KVM_X2APIC_API_USE_32BIT_IDS changes the behavior of KVM_SET_GSI_ROUTING, KVM_SIGNAL_MSI, KVM_SET_LAPIC, and KVM_GET_LAPIC, @@ -7814,6 +7816,27 @@ as a broadcast even in x2APIC mode in order to support physical x2APIC without interrupt remapping. This is undesirable in logical mode, where 0xff represents CPUs 0-7 in cluster 0. +Setting KVM_X2APIC_ENABLE_SUPPRESS_EOI_BROADCAST instructs KVM to enable +Suppress EOI Broadcasts. KVM will advertise support for Suppress EOI Broadcast +to the guest and suppress LAPIC EOI broadcasts when the guest sets the +Suppress EOI Broadcast bit in the SPIV register. + +Setting KVM_X2APIC_DISABLE_SUPPRESS_EOI_BROADCAST disables support for +Suppress EOI Broadcasts entirely, i.e. instructs KVM to NOT advertise support +to the guest. + +Modern VMMs should either enable KVM_X2APIC_ENABLE_SUPPRESS_EOI_BROADCAST or +KVM_X2APIC_DISABLE_SUPPRESS_EOI_BROADCAST. If not, legacy quirky behavior will +be used by KVM, which is to advertise support for Suppress EOI Broadcasts but +not actually suppressing EOI broadcasts. + +Currently, both KVM_X2APIC_ENABLE_SUPPRESS_EOI_BROADCAST and +KVM_X2APIC_DISABLE_SUPPRESS_EOI_BROADCAST must only be set when in split IRQCHIP +mode. Otherwise, the ioctl will fail with an EINVAL error. + +Setting both KVM_X2APIC_ENABLE_SUPPRESS_EOI_BROADCAST and +KVM_X2APIC_DISABLE_SUPPRESS_EOI_BROADCAST will fail with an EINVAL error. + 7.8 KVM_CAP_S390_USER_INSTR0 ---------------------------- diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h index 48598d017d6f..4a6d94dc7a2a 100644 --- a/arch/x86/include/asm/kvm_host.h +++ b/arch/x86/include/asm/kvm_host.h @@ -1229,6 +1229,12 @@ enum kvm_irqchip_mode { KVM_IRQCHIP_SPLIT, /* created with KVM_CAP_SPLIT_IRQCHIP */ }; +enum kvm_suppress_eoi_broadcast_mode { + KVM_SUPPRESS_EOI_BROADCAST_QUIRKED, /* Legacy behavior */ + KVM_SUPPRESS_EOI_BROADCAST_ENABLED, /* Enable Suppress EOI broadcast */ + KVM_SUPPRESS_EOI_BROADCAST_DISABLED /* Disable Suppress EOI broadcast */ +}; + struct kvm_x86_msr_filter { u8 count; bool default_allow:1; @@ -1480,6 +1486,7 @@ struct kvm_arch { bool x2apic_format; bool x2apic_broadcast_quirk_disabled; + enum kvm_suppress_eoi_broadcast_mode suppress_eoi_broadcast_mode; bool has_mapped_host_mmio; bool guest_can_read_msr_platform_info; diff --git a/arch/x86/include/uapi/asm/kvm.h b/arch/x86/include/uapi/asm/kvm.h index d420c9c066d4..d30241429fa8 100644 --- a/arch/x86/include/uapi/asm/kvm.h +++ b/arch/x86/include/uapi/asm/kvm.h @@ -913,8 +913,10 @@ struct kvm_sev_snp_launch_finish { __u64 pad1[4]; }; -#define KVM_X2APIC_API_USE_32BIT_IDS (1ULL << 0) -#define KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK (1ULL << 1) +#define KVM_X2APIC_API_USE_32BIT_IDS (_BITULL(0)) +#define KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK (_BITULL(1)) +#define KVM_X2APIC_ENABLE_SUPPRESS_EOI_BROADCAST (_BITULL(2)) +#define KVM_X2APIC_DISABLE_SUPPRESS_EOI_BROADCAST (_BITULL(3)) struct kvm_hyperv_eventfd { __u32 conn_id; diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c index 0ae7f913d782..1ef0bd3eff1e 100644 --- a/arch/x86/kvm/lapic.c +++ b/arch/x86/kvm/lapic.c @@ -105,6 +105,34 @@ bool kvm_apic_pending_eoi(struct kvm_vcpu *vcpu, int vector) apic_test_vector(vector, apic->regs + APIC_IRR); } +static inline bool kvm_lapic_advertise_suppress_eoi_broadcast(struct kvm *kvm) +{ + /* + * Advertise Suppress EOI broadcast support to the guest unless the VMM + * explicitly disabled it. + * + * Historically, KVM advertised this capability even though it did not + * actually suppress EOIs. + */ + return kvm->arch.suppress_eoi_broadcast_mode != + KVM_SUPPRESS_EOI_BROADCAST_DISABLED; +} + +static inline bool kvm_lapic_ignore_suppress_eoi_broadcast(struct kvm *kvm) +{ + /* + * Returns true if KVM should ignore the suppress EOI broadcast bit set by + * the guest and broadcast EOIs anyway. + * + * Only returns false when the VMM explicitly enabled Suppress EOI + * broadcast. If disabled by VMM, the bit should be ignored as it is not + * supported. Legacy behavior was to ignore the bit and broadcast EOIs + * anyway. + */ + return kvm->arch.suppress_eoi_broadcast_mode != + KVM_SUPPRESS_EOI_BROADCAST_ENABLED; +} + __read_mostly DEFINE_STATIC_KEY_FALSE(kvm_has_noapic_vcpu); EXPORT_SYMBOL_FOR_KVM_INTERNAL(kvm_has_noapic_vcpu); @@ -562,6 +590,7 @@ void kvm_apic_set_version(struct kvm_vcpu *vcpu) * IOAPIC. */ if (guest_cpu_cap_has(vcpu, X86_FEATURE_X2APIC) && + kvm_lapic_advertise_suppress_eoi_broadcast(vcpu->kvm) && !ioapic_in_kernel(vcpu->kvm)) v |= APIC_LVR_DIRECTED_EOI; kvm_lapic_set_reg(apic, APIC_LVR, v); @@ -1517,6 +1546,17 @@ static void kvm_ioapic_send_eoi(struct kvm_lapic *apic, int vector) /* Request a KVM exit to inform the userspace IOAPIC. */ if (irqchip_split(apic->vcpu->kvm)) { + /* + * Don't exit to userspace if the guest has enabled Directed + * EOI, a.k.a. Suppress EOI Broadcasts, in which case the local + * APIC doesn't broadcast EOIs (the guest must EOI the target + * I/O APIC(s) directly). Ignore the suppression if userspace + * has NOT explicitly enabled Suppress EOI broadcast. + */ + if ((kvm_lapic_get_reg(apic, APIC_SPIV) & APIC_SPIV_DIRECTED_EOI) && + !kvm_lapic_ignore_suppress_eoi_broadcast(apic->vcpu->kvm)) + return; + apic->vcpu->arch.pending_ioapic_eoi = vector; kvm_make_request(KVM_REQ_IOAPIC_EOI_EXIT, apic->vcpu); return; diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index c9c2aa6f4705..81b40fdb5f5f 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -121,8 +121,11 @@ static u64 __read_mostly efer_reserved_bits = ~((u64)EFER_SCE); #define KVM_CAP_PMU_VALID_MASK KVM_PMU_CAP_DISABLE -#define KVM_X2APIC_API_VALID_FLAGS (KVM_X2APIC_API_USE_32BIT_IDS | \ - KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK) +#define KVM_X2APIC_API_VALID_FLAGS \ + (KVM_X2APIC_API_USE_32BIT_IDS | \ + KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK | \ + KVM_X2APIC_ENABLE_SUPPRESS_EOI_BROADCAST | \ + KVM_X2APIC_DISABLE_SUPPRESS_EOI_BROADCAST) static void update_cr8_intercept(struct kvm_vcpu *vcpu); static void process_nmi(struct kvm_vcpu *vcpu); @@ -6777,12 +6780,22 @@ int kvm_vm_ioctl_enable_cap(struct kvm *kvm, r = -EINVAL; if (cap->args[0] & ~KVM_X2APIC_API_VALID_FLAGS) break; + if ((cap->args[0] & KVM_X2APIC_ENABLE_SUPPRESS_EOI_BROADCAST) && + (cap->args[0] & KVM_X2APIC_DISABLE_SUPPRESS_EOI_BROADCAST)) + break; + if (!irqchip_split(kvm) && + ((cap->args[0] & KVM_X2APIC_ENABLE_SUPPRESS_EOI_BROADCAST) || + (cap->args[0] & KVM_X2APIC_DISABLE_SUPPRESS_EOI_BROADCAST))) + break; if (cap->args[0] & KVM_X2APIC_API_USE_32BIT_IDS) kvm->arch.x2apic_format = true; if (cap->args[0] & KVM_X2APIC_API_DISABLE_BROADCAST_QUIRK) kvm->arch.x2apic_broadcast_quirk_disabled = true; - + if (cap->args[0] & KVM_X2APIC_ENABLE_SUPPRESS_EOI_BROADCAST) + kvm->arch.suppress_eoi_broadcast_mode = KVM_SUPPRESS_EOI_BROADCAST_ENABLED; + if (cap->args[0] & KVM_X2APIC_DISABLE_SUPPRESS_EOI_BROADCAST) + kvm->arch.suppress_eoi_broadcast_mode = KVM_SUPPRESS_EOI_BROADCAST_DISABLED; r = 0; break; case KVM_CAP_X86_DISABLE_EXITS: -- 2.39.3

1 week, 4 days

4
13
0 0

[PATCH 5.10.y v2 0/2] Backport 2 commits to fix a KASAN ext4 splat

by David Nyström

Backport commit:5701875f9609 ("ext4: fix out-of-bound read in ext4_xattr_inode_dec_ref_all()" to linux 5.10 branch. The fix depends on commit:69f3a3039b0d ("ext4: introduce ITAIL helper") In order to make a clean backport on stable kernel, backport 2 commits. It has a single merge conflict where static inline int, which changed to static int. Signed-off-by: David Nyström <david.nystrom(a)est.tech> --- Changes in v2: - Resend identical patchset with correct "Upstream commit" denotation. - Link to v1: https://patch.msgid.link/20251216-ext4_splat-v1-0-b76fd8748f44@est.tech --- Ye Bin (2): ext4: introduce ITAIL helper ext4: fix out-of-bound read in ext4_xattr_inode_dec_ref_all() fs/ext4/inode.c | 5 +++++ fs/ext4/xattr.c | 32 ++++---------------------------- fs/ext4/xattr.h | 10 ++++++++++ 3 files changed, 19 insertions(+), 28 deletions(-) --- base-commit: f964b940099f9982d723d4c77988d4b0dda9c165 change-id: 20251215-ext4_splat-f59c1acd9e88 Best regards, -- David Nyström <david.nystrom(a)est.tech>

1 week, 4 days

1
2
0 0

[PATCH v2,stable/linux-6.6.y] fbdev: Fix out-of-bounds issue in sys_fillrect()

by Gu Bowen

There was an out-of-bounds issue found by syzkaller test on v6.6. BUG: unable to handle page fault for address: ffffc90000c3f000 PGD 100000067 P4D 100000067 PUD 100c80067 PMD 10ac1c067 PTE 0 Oops: 0002 [#1] PREEMPT SMP KASAN PTI CPU: 3 PID: 6521 Comm: syz.3.1365 Not tainted 6.6.0+ #82 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.14.0-0-g155821a1990b-prebuilt.qemu.org 04/01/2014 RIP: 0010:memset64 arch/x86/include/asm/string_64.h:58 [inline] RIP: 0010:memset_l include/linux/string.h:168 [inline] RIP: 0010:bitfill_aligned drivers/video/fbdev/core/sysfillrect.c:53 [inline] RIP: 0010:bitfill_aligned+0x144/0x1c0 drivers/video/fbdev/core/sysfillrect.c:25 Code: 23 04 24 48 31 d0 49 89 46 f8 44 89 e0 44 29 f8 29 c3 e8 9f 39 49 fe 89 d8 31 d2 4c 89 f7 41 f7 f4 48 89 c3 48 89 c1 48 89 e8 <f3> 48 ab 31 ff 4c 89 ee e8 df 2f 49 fe 4d 85 ed 0f 84 6b ff ff ff RSP: 0018:ffff888119ce7418 EFLAGS: 00010246 RAX: 0000000000000000 RBX: 0000000000000180 RCX: 0000000000000180 RDX: 0000000000000000 RSI: ffffc90003873000 RDI: ffffc90000c3f000 RBP: 0000000000000000 R08: 0000000000006000 R09: 0000000000000040 R10: 0000000000000000 R11: 0000000000000001 R12: 0000000000000040 R13: 0000000000000000 R14: ffffc90000c3f000 R15: 0000000000000000 FS: 00007f1704b926c0(0000) GS:ffff8881f5980000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: ffffc90000c3f000 CR3: 00000001230d0002 CR4: 0000000000770ee0 DR0: 0000000000000000 DR1: 000000000000e000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 PKRU: 80000000 Call Trace: <TASK> sys_fillrect+0x429/0x830 drivers/video/fbdev/core/sysfillrect.c:281 drm_fbdev_generic_defio_fillrect+0x27/0x140 drivers/gpu/drm/drm_fbdev_generic.c:37 bit_clear+0x183/0x220 drivers/video/fbdev/core/bitblit.c:73 __fbcon_clear+0x5ea/0x670 drivers/video/fbdev/core/fbcon.c:1281 fbcon_scroll+0x41e/0x560 drivers/video/fbdev/core/fbcon.c:1847 con_scroll+0x464/0x6a0 drivers/tty/vt/vt.c:577 lf+0x274/0x2d0 drivers/tty/vt/vt.c:1461 do_con_trol+0x5ea/0x3d80 drivers/tty/vt/vt.c:2149 do_con_write+0x780/0x10c0 drivers/tty/vt/vt.c:2905 con_write+0x28/0xc0 drivers/tty/vt/vt.c:3245 do_output_char+0x5de/0x850 drivers/tty/n_tty.c:433 process_output drivers/tty/n_tty.c:500 [inline] n_tty_write+0x442/0xb00 drivers/tty/n_tty.c:2406 iterate_tty_write+0x2b5/0x630 drivers/tty/tty_io.c:1017 file_tty_write.constprop.0+0x20c/0x3b0 drivers/tty/tty_io.c:1088 call_write_iter include/linux/fs.h:2085 [inline] do_iter_readv_writev+0x210/0x3c0 fs/read_write.c:737 do_iter_write+0x181/0x4e0 fs/read_write.c:862 vfs_writev+0x15b/0x4d0 fs/read_write.c:935 do_writev+0x136/0x370 fs/read_write.c:978 do_syscall_x64 arch/x86/entry/common.c:51 [inline] do_syscall_64+0x59/0x110 arch/x86/entry/common.c:81 entry_SYSCALL_64_after_hwframe+0x78/0xe2 When the virtual console is rotated in the backend state, it can lead to inconsistencies between the size of the virtual console's size and its hook functions. In such cases, clearing the screen may result in out-of-bounds issue. This issue has already been fixed by commit eabb03293087 ("fbdev: Refactoring the fbcon packed pixel drawing routines") on v6.15-rc1, but it still exists in the stable version. It was unclear if there were other scenarios for this issue, and I tried to add pr_err in sys_fillrect. But as Helge pointed out[1], sys_fillrect is on the console printing code path and adding printing here might cause infinite loop. So let's just fix the known scenario. Fix it by moving set_blitting_type() to the visible area of the VC. Link: https://lore.kernel.org/all/aef7d5fd-2926-4c58-b720-4af58aa380d3@gmx.de/ [1] CC: stable(a)vger.kernel.org # for stable-6.6, fbdev had been refactored on 6.15-rc1 Fixes: 68648ed1f58d ("fbdev: add drawing functions for framebuffers in system RAM") Signed-off-by: Gu Bowen <gubowen5(a)huawei.com> --- drivers/video/fbdev/core/fbcon.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/video/fbdev/core/fbcon.c b/drivers/video/fbdev/core/fbcon.c index 78a5b22c8d15..8139ac8a666f 100644 --- a/drivers/video/fbdev/core/fbcon.c +++ b/drivers/video/fbdev/core/fbcon.c @@ -2702,9 +2702,9 @@ static void fbcon_modechanged(struct fb_info *info) return; p = &fb_display[vc->vc_num]; - set_blitting_type(vc, info); if (con_is_visible(vc)) { + set_blitting_type(vc, info); var_to_display(p, &info->var, info); cols = FBCON_SWAP(ops->rotate, info->var.xres, info->var.yres); rows = FBCON_SWAP(ops->rotate, info->var.yres, info->var.xres); -- 2.43.0

1 week, 4 days

2
1
0 0

[PATCH] rapidio: replace rio_free_net() with kfree() in rio_scan_alloc_net()

by Haoxiang Li

When idtab allocation fails, net is not registered with rio_add_net() yet, so kfree(net) is sufficient to release the memory. Fixes: e6b585ca6e81 ("rapidio: move net allocation into core code") Cc: stable(a)vger.kernel.org Signed-off-by: Haoxiang Li <lihaoxiang(a)isrc.iscas.ac.cn> --- drivers/rapidio/rio-scan.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/rapidio/rio-scan.c b/drivers/rapidio/rio-scan.c index c12941f71e2c..03615eb2d3d7 100644 --- a/drivers/rapidio/rio-scan.c +++ b/drivers/rapidio/rio-scan.c @@ -854,7 +854,7 @@ static struct rio_net *rio_scan_alloc_net(struct rio_mport *mport, if (idtab == NULL) { pr_err("RIO: failed to allocate destID table\n"); - rio_free_net(net); + kfree(net); net = NULL; } else { net->enum_data = idtab; -- 2.25.1

1 week, 4 days

1
0
0 0

[PATCH] pinctrl: qcom: lpass-lpi: mark the GPIO controller as sleeping

by Bartosz Golaszewski

From: Bartosz Golaszewski <bartosz.golaszewski(a)linaro.org> The gpio_chip settings in this driver say the controller can't sleep but it actually uses a mutex for synchronization. This triggers the following BUG(): [ 9.233659] BUG: sleeping function called from invalid context at kernel/locking/mutex.c:281 [ 9.233665] in_atomic(): 1, irqs_disabled(): 1, non_block: 0, pid: 554, name: (udev-worker) [ 9.233669] preempt_count: 1, expected: 0 [ 9.233673] RCU nest depth: 0, expected: 0 [ 9.233688] Tainted: [W]=WARN [ 9.233690] Hardware name: Dell Inc. Latitude 7455/0FK7MX, BIOS 2.10.1 05/20/2025 [ 9.233694] Call trace: [ 9.233696] show_stack+0x24/0x38 (C) [ 9.233709] dump_stack_lvl+0x40/0x88 [ 9.233716] dump_stack+0x18/0x24 [ 9.233722] __might_resched+0x148/0x160 [ 9.233731] __might_sleep+0x38/0x98 [ 9.233736] mutex_lock+0x30/0xd8 [ 9.233749] lpi_config_set+0x2e8/0x3c8 [pinctrl_lpass_lpi] [ 9.233757] lpi_gpio_direction_output+0x58/0x90 [pinctrl_lpass_lpi] [ 9.233761] gpiod_direction_output_raw_commit+0x110/0x428 [ 9.233772] gpiod_direction_output_nonotify+0x234/0x358 [ 9.233779] gpiod_direction_output+0x38/0xd0 [ 9.233786] gpio_shared_proxy_direction_output+0xb8/0x2a8 [gpio_shared_proxy] [ 9.233792] gpiod_direction_output_raw_commit+0x110/0x428 [ 9.233799] gpiod_direction_output_nonotify+0x234/0x358 [ 9.233806] gpiod_configure_flags+0x2c0/0x580 [ 9.233812] gpiod_find_and_request+0x358/0x4f8 [ 9.233819] gpiod_get_index+0x7c/0x98 [ 9.233826] devm_gpiod_get+0x34/0xb0 [ 9.233829] reset_gpio_probe+0x58/0x128 [reset_gpio] [ 9.233836] auxiliary_bus_probe+0xb0/0xf0 [ 9.233845] really_probe+0x14c/0x450 [ 9.233853] __driver_probe_device+0xb0/0x188 [ 9.233858] driver_probe_device+0x4c/0x250 [ 9.233863] __driver_attach+0xf8/0x2a0 [ 9.233868] bus_for_each_dev+0xf8/0x158 [ 9.233872] driver_attach+0x30/0x48 [ 9.233876] bus_add_driver+0x158/0x2b8 [ 9.233880] driver_register+0x74/0x118 [ 9.233886] __auxiliary_driver_register+0x94/0xe8 [ 9.233893] init_module+0x34/0xfd0 [reset_gpio] [ 9.233898] do_one_initcall+0xec/0x300 [ 9.233903] do_init_module+0x64/0x260 [ 9.233910] load_module+0x16c4/0x1900 [ 9.233915] __arm64_sys_finit_module+0x24c/0x378 [ 9.233919] invoke_syscall+0x4c/0xe8 [ 9.233925] el0_svc_common+0x8c/0xf0 [ 9.233929] do_el0_svc+0x28/0x40 [ 9.233934] el0_svc+0x38/0x100 [ 9.233938] el0t_64_sync_handler+0x84/0x130 [ 9.233943] el0t_64_sync+0x17c/0x180 Mark the controller as sleeping. Fixes: 6e261d1090d6 ("pinctrl: qcom: Add sm8250 lpass lpi pinctrl driver") Cc: stable(a)vger.kernel.org Reported-by: Val Packett <val(a)packett.cool> Closes: https://lore.kernel.org/all/98c0f185-b0e0-49ea-896c-f3972dd011ca@packett.co… Signed-off-by: Bartosz Golaszewski <bartosz.golaszewski(a)linaro.org> --- drivers/pinctrl/qcom/pinctrl-lpass-lpi.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/pinctrl/qcom/pinctrl-lpass-lpi.c b/drivers/pinctrl/qcom/pinctrl-lpass-lpi.c index 1c97ec44aa5ff..78212f9928430 100644 --- a/drivers/pinctrl/qcom/pinctrl-lpass-lpi.c +++ b/drivers/pinctrl/qcom/pinctrl-lpass-lpi.c @@ -498,7 +498,7 @@ int lpi_pinctrl_probe(struct platform_device *pdev) pctrl->chip.base = -1; pctrl->chip.ngpio = data->npins; pctrl->chip.label = dev_name(dev); - pctrl->chip.can_sleep = false; + pctrl->chip.can_sleep = true; mutex_init(&pctrl->lock); -- 2.51.0

1 week, 4 days

4
4
0 0

[PATCH] fbcon: Add check for return value

by Ваторопин Андрей

From: Andrey Vatoropin <a.vatoropin(a)crpt.ru> If fbcon_open() fails when called from con2fb_acquire_newinfo() then info->fbcon_par pointer remains NULL which is later dereferenced. Add check for return value of the function con2fb_acquire_newinfo() to avoid it. Found by Linux Verification Center (linuxtesting.org) with SVACE. Fixes: d1baa4ffa677 ("fbcon: set_con2fb_map fixes") Cc: stable(a)vger.kernel.org Signed-off-by: Andrey Vatoropin <a.vatoropin(a)crpt.ru> --- drivers/video/fbdev/core/fbcon.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/video/fbdev/core/fbcon.c b/drivers/video/fbdev/core/fbcon.c index e7e07eb2142e..7453377f3433 100644 --- a/drivers/video/fbdev/core/fbcon.c +++ b/drivers/video/fbdev/core/fbcon.c @@ -1047,7 +1047,8 @@ static void fbcon_init(struct vc_data *vc, bool init) return; if (!info->fbcon_par) - con2fb_acquire_newinfo(vc, info, vc->vc_num); + if (con2fb_acquire_newinfo(vc, info, vc->vc_num)) + return; /* If we are not the first console on this fb, copy the font from that console */ -- 2.43.0

1 week, 4 days

1
0
0 0

[PATCH] crypto: padlock-sha - Disable broken driver

by Eric Biggers

This driver is known broken, as it computes the wrong SHA-1 and SHA-256 hashes. Correctness needs to be the first priority for cryptographic code. Just disable it, allowing the standard (and actually correct) SHA-1 and SHA-256 implementations to take priority. Reported-by: larryw3i <larryw3i(a)yeah.net> Closes: https://lore.kernel.org/r/3af01fec-b4d3-4d0c-9450-2b722d4bbe39@yeah.net/ Closes: https://lists.debian.org/debian-kernel/2025/09/msg00019.html Closes: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1113996 Cc: stable(a)vger.kernel.org Cc: AlanSong-oc(a)zhaoxin.com Cc: CobeChen(a)zhaoxin.com Cc: GeorgeXue(a)zhaoxin.com Cc: HansHu(a)zhaoxin.com Cc: LeoLiu-oc(a)zhaoxin.com Cc: TonyWWang-oc(a)zhaoxin.com Cc: YunShen(a)zhaoxin.com Signed-off-by: Eric Biggers <ebiggers(a)kernel.org> --- This patch is targeting crypto/master drivers/crypto/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/crypto/Kconfig b/drivers/crypto/Kconfig index a6688d54984c..16ea3e741350 100644 --- a/drivers/crypto/Kconfig +++ b/drivers/crypto/Kconfig @@ -38,11 +38,11 @@ config CRYPTO_DEV_PADLOCK_AES If unsure say M. The compiled module will be called padlock-aes. config CRYPTO_DEV_PADLOCK_SHA tristate "PadLock driver for SHA1 and SHA256 algorithms" - depends on CRYPTO_DEV_PADLOCK + depends on CRYPTO_DEV_PADLOCK && BROKEN select CRYPTO_HASH select CRYPTO_SHA1 select CRYPTO_SHA256 help Use VIA PadLock for SHA1/SHA256 algorithms. base-commit: 59b0afd01b2ce353ab422ea9c8375b03db313a21 -- 2.51.2

1 week, 4 days

4
9
0 0

回复:6.12.50 regression: netem: cannot mix duplicating netems with other netems in tree.

by zyc zyc

Hello, Resend my last email without HTML. ---- zyc zyc <zyc199902(a)zohomail.cn> 在 Sat, 2025-11-29 18:57:01 写到：--- > Hello, maintainer > > I would like to report what appears to be a regression in 6.12.50 kernel release related to netem. > It rejects our configuration with the message: > Error: netem: cannot mix duplicating netems with other netems in tree. > > This breaks setups that previously worked correctly for many years. > > > Our team uses multiple netem qdiscs in the same HTB branch, arranged in a parallel fashion using a prio fan-out. Each branch of the prio qdisc has its own distinct netem instance with different duplication characteristics. > > This is used to emulate our production conditions where a single logical path fans out into two downstream segments, for example: > > two ECMP next hops with different misbehaviour characteristics, or > > > an HA firewall cluster where only one node is replaying frames, or > > > two LAG / ToR paths where one path intermittently duplicates packets. > > > In our environments, only a subset of flows are affected, and different downstream devices may cause different styles of duplication. > This regression breaks existing automated tests, training environments, and network simulation pipelines. > > I would be happy to provide our reproducer if needed. > > Thank you for your time and for maintaining Linux kernel. > > > > Best regards, > zyc > > >

1 week, 4 days

3
7
0 0

[PATCH 0/2] ASoC: SOF: ipc4-topology: fixes for 'exotic' format handling

by Peter Ujfalusi

Hi, The introduction of 8bit and FLOAT formats missed to cover the new corner cases they cause when the NHLT blobs are looked up. The two patch in this series fixes the 8bit and FLOAT format caused cases to be able to find the correct blob from NHLT. Regards, Peter --- Peter Ujfalusi (2): ASoC: SOF: ipc4-topology: Prefer 32-bit DMIC blobs for 8-bit formats as well ASoC: SOF: ipc4-topology: Convert FLOAT to S32 during blob selection sound/soc/sof/ipc4-topology.c | 24 +++++++++++++++--------- 1 file changed, 15 insertions(+), 9 deletions(-) -- 2.52.0

1 week, 4 days

2
3
0 0

[PATCH] nfc: llcp: stop processing on LLCP_CLOSED in nfc_llcp_recv_hdlc()

by Qianchang Zhao

nfc_llcp_sock_get() takes a reference on the LLCP socket via sock_hold(). In nfc_llcp_recv_hdlc(), the LLCP_CLOSED branch releases the socket lock and drops the reference, but the function continues to operate on llcp_sock/sk and later runs release_sock() and nfc_llcp_sock_put() again on the common exit path. Return immediately after the CLOSED cleanup to avoid refcount/lock imbalance and to avoid using the socket after dropping the reference. Reported-by: Qianchang Zhao <pioooooooooip(a)gmail.com> Reported-by: Zhitong Liu <liuzhitong1993(a)gmail.com> Cc: stable(a)vger.kernel.org Signed-off-by: Qianchang Zhao <pioooooooooip(a)gmail.com> --- net/nfc/llcp_core.c | 1 + 1 file changed, 1 insertion(+) diff --git a/net/nfc/llcp_core.c b/net/nfc/llcp_core.c index beeb3b4d2..be01ec9f4 100644 --- a/net/nfc/llcp_core.c +++ b/net/nfc/llcp_core.c @@ -1089,6 +1089,7 @@ static void nfc_llcp_recv_hdlc(struct nfc_llcp_local *local, if (sk->sk_state == LLCP_CLOSED) { release_sock(sk); nfc_llcp_sock_put(llcp_sock); + return; } /* Pass the payload upstream */ -- 2.34.1

1 week, 4 days

2
1
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror