December 2019 - Linux-stable-mirror

[PATCH bpf] bpf: cgroup: prevent out-of-order release of cgroup bpf

by Roman Gushchin

Before commit 4bfc0bb2c60e ("bpf: decouple the lifetime of cgroup_bpf from cgroup itself") cgroup bpf structures were released with corresponding cgroup structures. It guaranteed the hierarchical order of destruction: children were always first. It preserved attached programs from being released before their propagated copies. But with cgroup auto-detachment there are no such guarantees anymore: cgroup bpf is released as soon as the cgroup is offline and there are no live associated sockets. It means that an attached program can be detached and released, while its propagated copy is still living in the cgroup subtree. This will obviously lead to an use-after-free bug. To reproduce the issue the following script can be used: #!/bin/bash CGROOT=/sys/fs/cgroup mkdir -p ${CGROOT}/A ${CGROOT}/B ${CGROOT}/A/C sleep 1 ./test_cgrp2_attach ${CGROOT}/A egress & A_PID=$! ./test_cgrp2_attach ${CGROOT}/B egress & B_PID=$! echo $$ > ${CGROOT}/A/C/cgroup.procs iperf -s & S_PID=$! iperf -c localhost -t 100 & C_PID=$! sleep 1 echo $$ > ${CGROOT}/B/cgroup.procs echo ${S_PID} > ${CGROOT}/B/cgroup.procs echo ${C_PID} > ${CGROOT}/B/cgroup.procs sleep 1 rmdir ${CGROOT}/A/C rmdir ${CGROOT}/A sleep 1 kill -9 ${S_PID} ${C_PID} ${A_PID} ${B_PID} test_cgrp2_attach is an example from samples/bpf with the following patch applied (required to close cgroup and bpf program file descriptors after attachment): diff --git a/samples/bpf/test_cgrp2_attach.c b/samples/bpf/test_cgrp2_attach.c index 20fbd1241db3..7c7d0e91204d 100644 --- a/samples/bpf/test_cgrp2_attach.c +++ b/samples/bpf/test_cgrp2_attach.c @@ -111,6 +111,8 @@ static int attach_filter(int cg_fd, int type, int verdict) strerror(errno)); return EXIT_FAILURE; } + close(cg_fd); + close(prog_fd); while (1) { key = MAP_KEY_PACKETS; assert(bpf_map_lookup_elem(map_fd, &key, &pkt_cnt) == 0); On the unpatched kernel the following stacktrace can be obtained: [ 33.619799] BUG: unable to handle page fault for address: ffffbdb4801ab002 [ 33.620677] #PF: supervisor read access in kernel mode [ 33.621293] #PF: error_code(0x0000) - not-present page [ 33.621918] PGD 236d59067 P4D 236d59067 PUD 236d5c067 PMD 236d5d067 PTE 0 [ 33.622754] Oops: 0000 [#1] SMP NOPTI [ 33.623202] CPU: 0 PID: 601 Comm: iperf Not tainted 5.5.0-rc2+ #23 [ 33.623943] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS ?-20190727_073836-buildvm-ppc64le-16.ppc.f4 [ 33.625545] RIP: 0010:__cgroup_bpf_run_filter_skb+0x29f/0x3d0 [ 33.626231] Code: f6 0f 84 3a 01 00 00 49 8d 47 30 31 db 48 89 44 24 30 48 8b 45 08 65 48 89 05 4d 9d e0 64 48 8b d [ 33.628431] RSP: 0018:ffffbdb4802ffa90 EFLAGS: 00010246 [ 33.629051] RAX: 0000000000000000 RBX: 0000000000000000 RCX: 0000000000000034 [ 33.629906] RDX: 0000000000000000 RSI: ffff9ddf9d7a0000 RDI: ffff9ddf9b97f1c0 [ 33.630761] RBP: ffff9ddf9d4899d0 R08: ffff9ddfb67ddd80 R09: 0000000000010000 [ 33.631616] R10: 0000000000000070 R11: ffffbdb4802ffde8 R12: ffff9ddf9ba858e0 [ 33.632463] R13: 0000000000000001 R14: ffffbdb4801ab000 R15: ffff9ddf9ba858e0 [ 33.633306] FS: 00007f9d15ed9700(0000) GS:ffff9ddfb7c00000(0000) knlGS:0000000000000000 [ 33.634262] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 33.634945] CR2: ffffbdb4801ab002 CR3: 000000021b94e000 CR4: 00000000003406f0 [ 33.635809] Call Trace: [ 33.636118] ? __cgroup_bpf_run_filter_skb+0x2bf/0x3d0 [ 33.636728] ? __switch_to_asm+0x40/0x70 [ 33.637196] ip_finish_output+0x68/0xa0 [ 33.637654] ip_output+0x76/0xf0 [ 33.638046] ? __ip_finish_output+0x1c0/0x1c0 [ 33.638576] __ip_queue_xmit+0x157/0x410 [ 33.639049] __tcp_transmit_skb+0x535/0xaf0 [ 33.639557] tcp_write_xmit+0x378/0x1190 [ 33.640049] ? _copy_from_iter_full+0x8d/0x260 [ 33.640592] tcp_sendmsg_locked+0x2a2/0xdc0 [ 33.641098] ? sock_has_perm+0x10/0xa0 [ 33.641574] tcp_sendmsg+0x28/0x40 [ 33.641985] sock_sendmsg+0x57/0x60 [ 33.642411] sock_write_iter+0x97/0x100 [ 33.642876] new_sync_write+0x1b6/0x1d0 [ 33.643339] vfs_write+0xb6/0x1a0 [ 33.643752] ksys_write+0xa7/0xe0 [ 33.644156] do_syscall_64+0x5b/0x1b0 [ 33.644605] entry_SYSCALL_64_after_hwframe+0x44/0xa9 Fix this by grabbing a reference to the bpf structure of each ancestor on the initialization of the cgroup bpf structure, and dropping the reference at the end of releasing the cgroup bpf structure. This will restore the hierarchical order of cgroup bpf releasing, without adding any operations on hot paths. Thanks to Josef Bacik for the debugging and the initial analysis of the problem. Fixes: 4bfc0bb2c60e ("bpf: decouple the lifetime of cgroup_bpf from cgroup itself") Reported-by: Josef Bacik <josef(a)toxicpanda.com> Signed-off-by: Roman Gushchin <guro(a)fb.com> Cc: Alexei Starovoitov <ast(a)kernel.org> Cc: stable(a)vger.kernel.org --- kernel/bpf/cgroup.c | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-) diff --git a/kernel/bpf/cgroup.c b/kernel/bpf/cgroup.c index 4fb20ab179fe..9e43b72eb619 100644 --- a/kernel/bpf/cgroup.c +++ b/kernel/bpf/cgroup.c @@ -35,8 +35,8 @@ void cgroup_bpf_offline(struct cgroup *cgrp) */ static void cgroup_bpf_release(struct work_struct *work) { - struct cgroup *cgrp = container_of(work, struct cgroup, - bpf.release_work); + struct cgroup *p, *cgrp = container_of(work, struct cgroup, + bpf.release_work); enum bpf_cgroup_storage_type stype; struct bpf_prog_array *old_array; unsigned int type; @@ -65,6 +65,9 @@ static void cgroup_bpf_release(struct work_struct *work) mutex_unlock(&cgroup_mutex); + for (p = cgroup_parent(cgrp); p; p = cgroup_parent(p)) + cgroup_bpf_put(p); + percpu_ref_exit(&cgrp->bpf.refcnt); cgroup_put(cgrp); } @@ -199,6 +202,7 @@ int cgroup_bpf_inherit(struct cgroup *cgrp) */ #define NR ARRAY_SIZE(cgrp->bpf.effective) struct bpf_prog_array *arrays[NR] = {}; + struct cgroup *p; int ret, i; ret = percpu_ref_init(&cgrp->bpf.refcnt, cgroup_bpf_release_fn, 0, @@ -206,6 +210,9 @@ int cgroup_bpf_inherit(struct cgroup *cgrp) if (ret) return ret; + for (p = cgroup_parent(cgrp); p; p = cgroup_parent(p)) + cgroup_bpf_get(p); + for (i = 0; i < NR; i++) INIT_LIST_HEAD(&cgrp->bpf.progs[i]); -- 2.17.1

4 years, 4 months

3
8
0 0

[PATCH for 5.5 1/2] rseq: Fix: Clarify rseq.h UAPI rseq_cs memory reclaim requirements

by Mathieu Desnoyers

The rseq.h UAPI documents that the rseq_cs field must be cleared before reclaiming memory that contains the targeted struct rseq_cs. We should extend this comment to also dictate that the rseq_cs field must be cleared before reclaiming memory of the code pointed to by the rseq_cs start_ip and post_commit_offset fields. While we can expect that use of dlclose(3) will typically unmap both struct rseq_cs and its associated code at once, nothing would theoretically prevent a JIT from reclaiming the code without reclaiming the struct rseq_cs, which would erroneously allow the kernel to consider new code which is not a rseq critical section as a rseq critical section following a code reclaim. Suggested-by: Florian Weimer <fw(a)deneb.enyo.de> Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers(a)efficios.com> Cc: Florian Weimer <fw(a)deneb.enyo.de> Cc: Thomas Gleixner <tglx(a)linutronix.de> Cc: Peter Zijlstra (Intel) <peterz(a)infradead.org> Cc: "Paul E. McKenney" <paulmck(a)linux.ibm.com> Cc: Boqun Feng <boqun.feng(a)gmail.com> Cc: "H . Peter Anvin" <hpa(a)zytor.com> Cc: Paul Turner <pjt(a)google.com> Cc: Dmitry Vyukov <dvyukov(a)google.com> Cc: Neel Natu <neelnatu(a)google.com> Cc: linux-api(a)vger.kernel.org --- include/uapi/linux/rseq.h | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/include/uapi/linux/rseq.h b/include/uapi/linux/rseq.h index 9a402fdb60e9..6f26b0b148a6 100644 --- a/include/uapi/linux/rseq.h +++ b/include/uapi/linux/rseq.h @@ -100,7 +100,9 @@ struct rseq { * instruction sequence block, as well as when the kernel detects that * it is preempting or delivering a signal outside of the range * targeted by the rseq_cs. Also needs to be set to NULL by user-space - * before reclaiming memory that contains the targeted struct rseq_cs. + * before reclaiming memory that contains the targeted struct rseq_cs + * or reclaiming memory that contains the code refered to by the + * start_ip and post_commit_offset fields of struct rseq_cs. * * Read and set by the kernel. Set by user-space with single-copy * atomicity semantics. This field should only be updated by the -- 2.17.1

4 years, 4 months

3
10
0 0

[PATCH v2 13/13] KVM: x86: Protect pmu_intel.c from Spectre-v1/L1TF attacks

by Marios Pomonis

This fixes Spectre-v1/L1TF vulnerabilities in intel_find_fixed_event() and intel_rdpmc_ecx_to_pmc(). kvm_rdpmc() (ancestor of intel_find_fixed_event()) and reprogram_fixed_counter() (ancestor of intel_rdpmc_ecx_to_pmc()) are exported symbols so KVM should treat them conservatively from a security perspective. Fixes: commit 25462f7f5295 ("KVM: x86/vPMU: Define kvm_pmu_ops to support vPMU function dispatch") Signed-off-by: Nick Finco <nifi(a)google.com> Signed-off-by: Marios Pomonis <pomonis(a)google.com> Reviewed-by: Andrew Honig <ahonig(a)google.com> Cc: stable(a)vger.kernel.org --- arch/x86/kvm/vmx/pmu_intel.c | 24 ++++++++++++++++-------- 1 file changed, 16 insertions(+), 8 deletions(-) diff --git a/arch/x86/kvm/vmx/pmu_intel.c b/arch/x86/kvm/vmx/pmu_intel.c index 7023138b1cb0..34a3a17bb6d7 100644 --- a/arch/x86/kvm/vmx/pmu_intel.c +++ b/arch/x86/kvm/vmx/pmu_intel.c @@ -86,10 +86,14 @@ static unsigned intel_find_arch_event(struct kvm_pmu *pmu, static unsigned intel_find_fixed_event(int idx) { - if (idx >= ARRAY_SIZE(fixed_pmc_events)) + u32 event; + size_t size = ARRAY_SIZE(fixed_pmc_events); + + if (idx >= size) return PERF_COUNT_HW_MAX; - return intel_arch_events[fixed_pmc_events[idx]].event_type; + event = fixed_pmc_events[array_index_nospec(idx, size)]; + return intel_arch_events[event].event_type; } /* check if a PMC is enabled by comparing it with globl_ctrl bits. */ @@ -130,16 +134,20 @@ static struct kvm_pmc *intel_rdpmc_ecx_to_pmc(struct kvm_vcpu *vcpu, struct kvm_pmu *pmu = vcpu_to_pmu(vcpu); bool fixed = idx & (1u << 30); struct kvm_pmc *counters; + unsigned int num_counters; idx &= ~(3u << 30); - if (!fixed && idx >= pmu->nr_arch_gp_counters) - return NULL; - if (fixed && idx >= pmu->nr_arch_fixed_counters) + if (fixed) { + counters = pmu->fixed_counters; + num_counters = pmu->nr_arch_fixed_counters; + } else { + counters = pmu->gp_counters; + num_counters = pmu->nr_arch_gp_counters; + } + if (idx >= num_counters) return NULL; - counters = fixed ? pmu->fixed_counters : pmu->gp_counters; *mask &= pmu->counter_bitmask[fixed ? KVM_PMC_FIXED : KVM_PMC_GP]; - - return &counters[idx]; + return &counters[array_index_nospec(idx, num_counters)]; } static bool intel_is_valid_msr(struct kvm_vcpu *vcpu, u32 msr) -- 2.24.0.525.g8f36a354ae-goog

4 years, 4 months

2
1
0 0

[PATCH v2 12/13] KVM: x86: Protect DR-based index computations from Spectre-v1/L1TF attacks

by Marios Pomonis

This fixes a Spectre-v1/L1TF vulnerability in __kvm_set_dr() and kvm_get_dr(). Both kvm_get_dr() and kvm_set_dr() (a wrapper of __kvm_set_dr()) are exported symbols so KVM should tream them conservatively from a security perspective. Fixes: commit 020df0794f57 ("KVM: move DR register access handling into generic code") Signed-off-by: Nick Finco <nifi(a)google.com> Signed-off-by: Marios Pomonis <pomonis(a)google.com> Reviewed-by: Andrew Honig <ahonig(a)google.com> Cc: stable(a)vger.kernel.org --- arch/x86/kvm/x86.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index a9e66f09422e..9a2789652231 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -1057,9 +1057,11 @@ static u64 kvm_dr6_fixed(struct kvm_vcpu *vcpu) static int __kvm_set_dr(struct kvm_vcpu *vcpu, int dr, unsigned long val) { + size_t size = ARRAY_SIZE(vcpu->arch.db); + switch (dr) { case 0 ... 3: - vcpu->arch.db[dr] = val; + vcpu->arch.db[array_index_nospec(dr, size)] = val; if (!(vcpu->guest_debug & KVM_GUESTDBG_USE_HW_BP)) vcpu->arch.eff_db[dr] = val; break; @@ -1096,9 +1098,11 @@ EXPORT_SYMBOL_GPL(kvm_set_dr); int kvm_get_dr(struct kvm_vcpu *vcpu, int dr, unsigned long *val) { + size_t size = ARRAY_SIZE(vcpu->arch.db); + switch (dr) { case 0 ... 3: - *val = vcpu->arch.db[dr]; + *val = vcpu->arch.db[array_index_nospec(dr, size)]; break; case 4: /* fall through */ -- 2.24.0.525.g8f36a354ae-goog

4 years, 4 months

2
1
0 0

[PATCH v2 09/13] KVM: x86: Protect MSR-based index computations from Spectre-v1/L1TF attacks in x86.c

by Marios Pomonis

This fixes a Spectre-v1/L1TF vulnerability in set_msr_mce() and get_msr_mce(). Both functions contain index computations based on the (attacker-controlled) MSR number. Fixes: commit 890ca9aefa78 ("KVM: Add MCE support") Signed-off-by: Nick Finco <nifi(a)google.com> Signed-off-by: Marios Pomonis <pomonis(a)google.com> Reviewed-by: Andrew Honig <ahonig(a)google.com> Cc: stable(a)vger.kernel.org --- arch/x86/kvm/x86.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index a256e09f321a..a9e66f09422e 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -2496,7 +2496,10 @@ static int set_msr_mce(struct kvm_vcpu *vcpu, struct msr_data *msr_info) default: if (msr >= MSR_IA32_MC0_CTL && msr < MSR_IA32_MCx_CTL(bank_num)) { - u32 offset = msr - MSR_IA32_MC0_CTL; + u32 offset = array_index_nospec( + msr - MSR_IA32_MC0_CTL, + MSR_IA32_MCx_CTL(bank_num) - MSR_IA32_MC0_CTL); + /* only 0 or all 1s can be written to IA32_MCi_CTL * some Linux kernels though clear bit 10 in bank 4 to * workaround a BIOS/GART TBL issue on AMD K8s, ignore @@ -2937,7 +2940,10 @@ static int get_msr_mce(struct kvm_vcpu *vcpu, u32 msr, u64 *pdata, bool host) default: if (msr >= MSR_IA32_MC0_CTL && msr < MSR_IA32_MCx_CTL(bank_num)) { - u32 offset = msr - MSR_IA32_MC0_CTL; + u32 offset = array_index_nospec( + msr - MSR_IA32_MC0_CTL, + MSR_IA32_MCx_CTL(bank_num) - MSR_IA32_MC0_CTL); + data = vcpu->arch.mce_banks[offset]; break; } -- 2.24.0.525.g8f36a354ae-goog

4 years, 4 months

2
1
0 0

[PATCH v2 08/13] KVM: x86: Protect MSR-based index computations in pmu.h from Spectre-v1/L1TF attacks

by Marios Pomonis

This fixes a Spectre-v1/L1TF vulnerability in the get_gp_pmc() and get_fixed_pmc() functions. They both contain index computations based on the (attacker-controlled) MSR number. Fixes: commit 25462f7f5295 ("KVM: x86/vPMU: Define kvm_pmu_ops to support vPMU function dispatch") Signed-off-by: Nick Finco <nifi(a)google.com> Signed-off-by: Marios Pomonis <pomonis(a)google.com> Reviewed-by: Andrew Honig <ahonig(a)google.com> Cc: stable(a)vger.kernel.org --- arch/x86/kvm/pmu.h | 18 ++++++++++++++---- 1 file changed, 14 insertions(+), 4 deletions(-) diff --git a/arch/x86/kvm/pmu.h b/arch/x86/kvm/pmu.h index 7ebb62326c14..13332984b6d5 100644 --- a/arch/x86/kvm/pmu.h +++ b/arch/x86/kvm/pmu.h @@ -2,6 +2,8 @@ #ifndef __KVM_X86_PMU_H #define __KVM_X86_PMU_H +#include <linux/nospec.h> + #define vcpu_to_pmu(vcpu) (&(vcpu)->arch.pmu) #define pmu_to_vcpu(pmu) (container_of((pmu), struct kvm_vcpu, arch.pmu)) #define pmc_to_pmu(pmc) (&(pmc)->vcpu->arch.pmu) @@ -102,8 +104,12 @@ static inline bool kvm_valid_perf_global_ctrl(struct kvm_pmu *pmu, static inline struct kvm_pmc *get_gp_pmc(struct kvm_pmu *pmu, u32 msr, u32 base) { - if (msr >= base && msr < base + pmu->nr_arch_gp_counters) - return &pmu->gp_counters[msr - base]; + if (msr >= base && msr < base + pmu->nr_arch_gp_counters) { + u32 index = array_index_nospec(msr - base, + pmu->nr_arch_gp_counters); + + return &pmu->gp_counters[index]; + } return NULL; } @@ -113,8 +119,12 @@ static inline struct kvm_pmc *get_fixed_pmc(struct kvm_pmu *pmu, u32 msr) { int base = MSR_CORE_PERF_FIXED_CTR0; - if (msr >= base && msr < base + pmu->nr_arch_fixed_counters) - return &pmu->fixed_counters[msr - base]; + if (msr >= base && msr < base + pmu->nr_arch_fixed_counters) { + u32 index = array_index_nospec(msr - base, + pmu->nr_arch_fixed_counters); + + return &pmu->fixed_counters[index]; + } return NULL; } -- 2.24.0.525.g8f36a354ae-goog

4 years, 4 months

2
1
0 0

[PATCH v2 07/13] KVM: x86: Protect MSR-based index computations in fixed_msr_to_seg_unit() from Spectre-v1/L1TF attacks

by Marios Pomonis

This fixes a Spectre-v1/L1TF vulnerability in fixed_msr_to_seg_unit(). This function contains index computations based on the (attacker-controlled) MSR number. Fixes: commit de9aef5e1ad6 ("KVM: MTRR: introduce fixed_mtrr_segment table") Signed-off-by: Nick Finco <nifi(a)google.com> Signed-off-by: Marios Pomonis <pomonis(a)google.com> Reviewed-by: Andrew Honig <ahonig(a)google.com> Cc: stable(a)vger.kernel.org --- arch/x86/kvm/mtrr.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/arch/x86/kvm/mtrr.c b/arch/x86/kvm/mtrr.c index 25ce3edd1872..7f0059aa30e1 100644 --- a/arch/x86/kvm/mtrr.c +++ b/arch/x86/kvm/mtrr.c @@ -192,11 +192,15 @@ static bool fixed_msr_to_seg_unit(u32 msr, int *seg, int *unit) break; case MSR_MTRRfix16K_80000 ... MSR_MTRRfix16K_A0000: *seg = 1; - *unit = msr - MSR_MTRRfix16K_80000; + *unit = array_index_nospec( + msr - MSR_MTRRfix16K_80000, + MSR_MTRRfix16K_A0000 - MSR_MTRRfix16K_80000 + 1); break; case MSR_MTRRfix4K_C0000 ... MSR_MTRRfix4K_F8000: *seg = 2; - *unit = msr - MSR_MTRRfix4K_C0000; + *unit = array_index_nospec( + msr - MSR_MTRRfix4K_C0000, + MSR_MTRRfix4K_F8000 - MSR_MTRRfix4K_C0000 + 1); break; default: return false; -- 2.24.0.525.g8f36a354ae-goog

4 years, 4 months

2
1
0 0

[PATCH v2 06/13] KVM: x86: Protect kvm_lapic_reg_write() from Spectre-v1/L1TF attacks

by Marios Pomonis

This fixes a Spectre-v1/L1TF vulnerability in kvm_lapic_reg_write(). This function contains index computations based on the (attacker-controlled) MSR number. Fixes: commit 0105d1a52640 ("KVM: x2apic interface to lapic") Signed-off-by: Nick Finco <nifi(a)google.com> Signed-off-by: Marios Pomonis <pomonis(a)google.com> Reviewed-by: Andrew Honig <ahonig(a)google.com> Cc: stable(a)vger.kernel.org --- arch/x86/kvm/lapic.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c index cf9177b4a07f..3323115f52d5 100644 --- a/arch/x86/kvm/lapic.c +++ b/arch/x86/kvm/lapic.c @@ -1963,15 +1963,20 @@ int kvm_lapic_reg_write(struct kvm_lapic *apic, u32 reg, u32 val) case APIC_LVTTHMR: case APIC_LVTPC: case APIC_LVT1: - case APIC_LVTERR: + case APIC_LVTERR: { /* TODO: Check vector */ + size_t size; + u32 index; + if (!kvm_apic_sw_enabled(apic)) val |= APIC_LVT_MASKED; - - val &= apic_lvt_mask[(reg - APIC_LVTT) >> 4]; + size = ARRAY_SIZE(apic_lvt_mask); + index = array_index_nospec( + (reg - APIC_LVTT) >> 4, size); + val &= apic_lvt_mask[index]; kvm_lapic_set_reg(apic, reg, val); - break; + } case APIC_LVTT: if (!kvm_apic_sw_enabled(apic)) -- 2.24.0.525.g8f36a354ae-goog

4 years, 4 months

2
1
0 0

[PATCH v2 05/13] KVM: x86: Protect ioapic_write_indirect() from Spectre-v1/L1TF attacks

by Marios Pomonis

This fixes a Spectre-v1/L1TF vulnerability in ioapic_write_indirect(). This function contains index computations based on the (attacker-controlled) IOREGSEL register. This patch depends on patch "KVM: x86: Protect ioapic_read_indirect() from Spectre-v1/L1TF attacks". Fixes: commit 70f93dae32ac ("KVM: Use temporary variable to shorten lines.") Signed-off-by: Nick Finco <nifi(a)google.com> Signed-off-by: Marios Pomonis <pomonis(a)google.com> Reviewed-by: Andrew Honig <ahonig(a)google.com> Cc: stable(a)vger.kernel.org --- arch/x86/kvm/ioapic.c | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/x86/kvm/ioapic.c b/arch/x86/kvm/ioapic.c index 0c672eefaabe..8aa58727045e 100644 --- a/arch/x86/kvm/ioapic.c +++ b/arch/x86/kvm/ioapic.c @@ -294,6 +294,7 @@ static void ioapic_write_indirect(struct kvm_ioapic *ioapic, u32 val) if (index >= IOAPIC_NUM_PINS) return; + index = array_index_nospec(index, IOAPIC_NUM_PINS); e = &ioapic->redirtbl[index]; mask_before = e->fields.mask; /* Preserve read-only fields */ -- 2.24.0.525.g8f36a354ae-goog

4 years, 4 months

2
1
0 0

[PATCH v2 04/13] KVM: x86: Protect ioapic_read_indirect() from Spectre-v1/L1TF attacks

by Marios Pomonis

This fixes a Spectre-v1/L1TF vulnerability in ioapic_read_indirect(). This function contains index computations based on the (attacker-controlled) IOREGSEL register. Fixes: commit a2c118bfab8b ("KVM: Fix bounds checking in ioapic indirect register reads (CVE-2013-1798)") Signed-off-by: Nick Finco <nifi(a)google.com> Signed-off-by: Marios Pomonis <pomonis(a)google.com> Reviewed-by: Andrew Honig <ahonig(a)google.com> Cc: stable(a)vger.kernel.org --- arch/x86/kvm/ioapic.c | 14 ++++++++------ 1 file changed, 8 insertions(+), 6 deletions(-) diff --git a/arch/x86/kvm/ioapic.c b/arch/x86/kvm/ioapic.c index 9fd2dd89a1c5..0c672eefaabe 100644 --- a/arch/x86/kvm/ioapic.c +++ b/arch/x86/kvm/ioapic.c @@ -36,6 +36,7 @@ #include <linux/io.h> #include <linux/slab.h> #include <linux/export.h> +#include <linux/nospec.h> #include <asm/processor.h> #include <asm/page.h> #include <asm/current.h> @@ -68,13 +69,14 @@ static unsigned long ioapic_read_indirect(struct kvm_ioapic *ioapic, default: { u32 redir_index = (ioapic->ioregsel - 0x10) >> 1; - u64 redir_content; + u64 redir_content = ~0ULL; - if (redir_index < IOAPIC_NUM_PINS) - redir_content = - ioapic->redirtbl[redir_index].bits; - else - redir_content = ~0ULL; + if (redir_index < IOAPIC_NUM_PINS) { + u32 index = array_index_nospec( + redir_index, IOAPIC_NUM_PINS); + + redir_content = ioapic->redirtbl[index].bits; + } result = (ioapic->ioregsel & 0x1) ? (redir_content >> 32) & 0xffffffff : -- 2.24.0.525.g8f36a354ae-goog

4 years, 4 months

2
1
0 0

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror December 2019