From e2bc70c7d92e31116f3fb14f9e8e68fd2ad6c58f Mon Sep 17 00:00:00 2001 From: CIQ Kernel Automation Date: Mon, 29 Jun 2026 08:02:43 +0000 Subject: [PATCH 1/9] fs: sysfs: Fix reference leak in sysfs_break_active_protection() jira VULN-980 cve CVE-2024-26993 commit-author Alan Stern commit a90bca2228c0646fc29a72689d308e5fe03e6d78 The sysfs_break_active_protection() routine has an obvious reference leak in its error path. If the call to kernfs_find_and_get() fails then kn will be NULL, so the companion sysfs_unbreak_active_protection() routine won't get called (and would only cause an access violation by trying to dereference kn->parent if it was called). As a result, the reference to kobj acquired at the start of the function will never be released. Fix the leak by adding an explicit kobject_put() call when kn is NULL. Signed-off-by: Alan Stern Fixes: 2afc9166f79b ("scsi: sysfs: Introduce sysfs_{un,}break_active_protection()") Cc: Bart Van Assche Cc: stable@vger.kernel.org Reviewed-by: Bart Van Assche Acked-by: Tejun Heo Link: https://lore.kernel.org/r/8a4d3f0f-c5e3-4b70-a188-0ca433f9e6f9@rowland.harvard.edu Signed-off-by: Greg Kroah-Hartman (cherry picked from commit a90bca2228c0646fc29a72689d308e5fe03e6d78) Signed-off-by: CIQ Kernel Automation --- fs/sysfs/file.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/fs/sysfs/file.c b/fs/sysfs/file.c index 9aefa7779b29a..2a52f9df0b27f 100644 --- a/fs/sysfs/file.c +++ b/fs/sysfs/file.c @@ -440,6 +440,8 @@ struct kernfs_node *sysfs_break_active_protection(struct kobject *kobj, kn = kernfs_find_and_get(kobj->sd, attr->name); if (kn) kernfs_break_active_protection(kn); + else + kobject_put(kobj); return kn; } EXPORT_SYMBOL_GPL(sysfs_break_active_protection); From b7fc1a453e002ccd3c1a6a36b92a5f16a74e0c20 Mon Sep 17 00:00:00 2001 From: CIQ Kernel Automation Date: Mon, 29 Jun 2026 08:03:37 +0000 Subject: [PATCH 2/9] cpufreq: exit() callback is optional jira VULN-5516 cve CVE-2024-38615 commit-author Viresh Kumar commit b8f85833c05730d631576008daaa34096bc7f3ce The exit() callback is optional and shouldn't be called without checking a valid pointer first. Also, we must clear freq_table pointer even if the exit() callback isn't present. Signed-off-by: Viresh Kumar Fixes: 91a12e91dc39 ("cpufreq: Allow light-weight tear down and bring up of CPUs") Fixes: f339f3541701 ("cpufreq: Rearrange locking in cpufreq_remove_dev()") Reported-by: Lizhe Signed-off-by: Rafael J. Wysocki (cherry picked from commit b8f85833c05730d631576008daaa34096bc7f3ce) Signed-off-by: CIQ Kernel Automation --- drivers/cpufreq/cpufreq.c | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c index afd50b6927f6c..0f62cb814e514 100644 --- a/drivers/cpufreq/cpufreq.c +++ b/drivers/cpufreq/cpufreq.c @@ -1623,10 +1623,13 @@ static void __cpufreq_offline(unsigned int cpu, struct cpufreq_policy *policy) */ if (cpufreq_driver->offline) { cpufreq_driver->offline(policy); - } else if (cpufreq_driver->exit) { - cpufreq_driver->exit(policy); - policy->freq_table = NULL; + return; } + + if (cpufreq_driver->exit) + cpufreq_driver->exit(policy); + + policy->freq_table = NULL; } static int cpufreq_offline(unsigned int cpu) @@ -1675,7 +1678,7 @@ static void cpufreq_remove_dev(struct device *dev, struct subsys_interface *sif) } /* We did light-weight exit earlier, do full tear down now */ - if (cpufreq_driver->offline) + if (cpufreq_driver->offline && cpufreq_driver->exit) cpufreq_driver->exit(policy); up_write(&policy->rwsem); From 2a1cde6a144d166d5e85a78b23ac3a5b9551ba95 Mon Sep 17 00:00:00 2001 From: CIQ Kernel Automation Date: Mon, 29 Jun 2026 08:04:18 +0000 Subject: [PATCH 3/9] nouveau: fix instmem race condition around ptr stores jira VULN-37531 cve CVE-2024-26984 commit-author Dave Airlie commit fff1386cc889d8fb4089d285f883f8cba62d82ce Running a lot of VK CTS in parallel against nouveau, once every few hours you might see something like this crash. BUG: kernel NULL pointer dereference, address: 0000000000000008 PGD 8000000114e6e067 P4D 8000000114e6e067 PUD 109046067 PMD 0 Oops: 0000 [#1] PREEMPT SMP PTI CPU: 7 PID: 53891 Comm: deqp-vk Not tainted 6.8.0-rc6+ #27 Hardware name: Gigabyte Technology Co., Ltd. Z390 I AORUS PRO WIFI/Z390 I AORUS PRO WIFI-CF, BIOS F8 11/05/2021 RIP: 0010:gp100_vmm_pgt_mem+0xe3/0x180 [nouveau] Code: c7 48 01 c8 49 89 45 58 85 d2 0f 84 95 00 00 00 41 0f b7 46 12 49 8b 7e 08 89 da 42 8d 2c f8 48 8b 47 08 41 83 c7 01 48 89 ee <48> 8b 40 08 ff d0 0f 1f 00 49 8b 7e 08 48 89 d9 48 8d 75 04 48 c1 RSP: 0000:ffffac20c5857838 EFLAGS: 00010202 RAX: 0000000000000000 RBX: 00000000004d8001 RCX: 0000000000000001 RDX: 00000000004d8001 RSI: 00000000000006d8 RDI: ffffa07afe332180 RBP: 00000000000006d8 R08: ffffac20c5857ad0 R09: 0000000000ffff10 R10: 0000000000000001 R11: ffffa07af27e2de0 R12: 000000000000001c R13: ffffac20c5857ad0 R14: ffffa07a96fe9040 R15: 000000000000001c FS: 00007fe395eed7c0(0000) GS:ffffa07e2c980000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000000000000008 CR3: 000000011febe001 CR4: 00000000003706f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: ... ? gp100_vmm_pgt_mem+0xe3/0x180 [nouveau] ? gp100_vmm_pgt_mem+0x37/0x180 [nouveau] nvkm_vmm_iter+0x351/0xa20 [nouveau] ? __pfx_nvkm_vmm_ref_ptes+0x10/0x10 [nouveau] ? __pfx_gp100_vmm_pgt_mem+0x10/0x10 [nouveau] ? __pfx_gp100_vmm_pgt_mem+0x10/0x10 [nouveau] ? __lock_acquire+0x3ed/0x2170 ? __pfx_gp100_vmm_pgt_mem+0x10/0x10 [nouveau] nvkm_vmm_ptes_get_map+0xc2/0x100 [nouveau] ? __pfx_nvkm_vmm_ref_ptes+0x10/0x10 [nouveau] ? __pfx_gp100_vmm_pgt_mem+0x10/0x10 [nouveau] nvkm_vmm_map_locked+0x224/0x3a0 [nouveau] Adding any sort of useful debug usually makes it go away, so I hand wrote the function in a line, and debugged the asm. Every so often pt->memory->ptrs is NULL. This ptrs ptr is set in the nv50_instobj_acquire called from nvkm_kmap. If Thread A and Thread B both get to nv50_instobj_acquire around the same time, and Thread A hits the refcount_set line, and in lockstep thread B succeeds at refcount_inc_not_zero, there is a chance the ptrs value won't have been stored since refcount_set is unordered. Force a memory barrier here, I picked smp_mb, since we want it on all CPUs and it's write followed by a read. v2: use paired smp_rmb/smp_wmb. Cc: Fixes: be55287aa5ba ("drm/nouveau/imem/nv50: embed nvkm_instobj directly into nv04_instobj") Signed-off-by: Dave Airlie Signed-off-by: Danilo Krummrich Link: https://patchwork.freedesktop.org/patch/msgid/20240411011510.2546857-1-airlied@gmail.com (cherry picked from commit fff1386cc889d8fb4089d285f883f8cba62d82ce) Signed-off-by: CIQ Kernel Automation --- drivers/gpu/drm/nouveau/nvkm/subdev/instmem/nv50.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/instmem/nv50.c b/drivers/gpu/drm/nouveau/nvkm/subdev/instmem/nv50.c index c51bac76174c1..9fe5b6a36ab98 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/instmem/nv50.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/instmem/nv50.c @@ -221,8 +221,11 @@ nv50_instobj_acquire(struct nvkm_memory *memory) void __iomem *map = NULL; /* Already mapped? */ - if (refcount_inc_not_zero(&iobj->maps)) + if (refcount_inc_not_zero(&iobj->maps)) { + /* read barrier match the wmb on refcount set */ + smp_rmb(); return iobj->map; + } /* Take the lock, and re-check that another thread hasn't * already mapped the object in the meantime. @@ -249,6 +252,8 @@ nv50_instobj_acquire(struct nvkm_memory *memory) iobj->base.memory.ptrs = &nv50_instobj_fast; else iobj->base.memory.ptrs = &nv50_instobj_slow; + /* barrier to ensure the ptrs are written before refcount is set */ + smp_wmb(); refcount_set(&iobj->maps, 1); } From 25df59ad9417a76d0193480c0bf4656e43266d00 Mon Sep 17 00:00:00 2001 From: CIQ Kernel Automation Date: Mon, 29 Jun 2026 08:05:02 +0000 Subject: [PATCH 4/9] cppc_cpufreq: Fix possible null pointer dereference jira VULN-5484 cve CVE-2024-38573 commit-author Aleksandr Mishin commit cf7de25878a1f4508c69dc9f6819c21ba177dbfe cppc_cpufreq_get_rate() and hisi_cppc_cpufreq_get_rate() can be called from different places with various parameters. So cpufreq_cpu_get() can return null as 'policy' in some circumstances. Fix this bug by adding null return check. Found by Linux Verification Center (linuxtesting.org) with SVACE. Fixes: a28b2bfc099c ("cppc_cpufreq: replace per-cpu data array with a list") Signed-off-by: Aleksandr Mishin Signed-off-by: Viresh Kumar (cherry picked from commit cf7de25878a1f4508c69dc9f6819c21ba177dbfe) Signed-off-by: CIQ Kernel Automation --- drivers/cpufreq/cppc_cpufreq.c | 14 ++++++++++++-- 1 file changed, 12 insertions(+), 2 deletions(-) diff --git a/drivers/cpufreq/cppc_cpufreq.c b/drivers/cpufreq/cppc_cpufreq.c index 432dfb4e8027e..68109bcab89e6 100644 --- a/drivers/cpufreq/cppc_cpufreq.c +++ b/drivers/cpufreq/cppc_cpufreq.c @@ -840,10 +840,15 @@ static unsigned int cppc_cpufreq_get_rate(unsigned int cpu) { struct cppc_perf_fb_ctrs fb_ctrs_t0 = {0}, fb_ctrs_t1 = {0}; struct cpufreq_policy *policy = cpufreq_cpu_get(cpu); - struct cppc_cpudata *cpu_data = policy->driver_data; + struct cppc_cpudata *cpu_data; u64 delivered_perf; int ret; + if (!policy) + return -ENODEV; + + cpu_data = policy->driver_data; + cpufreq_cpu_put(policy); ret = cppc_get_perf_ctrs(cpu, &fb_ctrs_t0); @@ -923,10 +928,15 @@ static struct cpufreq_driver cppc_cpufreq_driver = { static unsigned int hisi_cppc_cpufreq_get_rate(unsigned int cpu) { struct cpufreq_policy *policy = cpufreq_cpu_get(cpu); - struct cppc_cpudata *cpu_data = policy->driver_data; + struct cppc_cpudata *cpu_data; u64 desired_perf; int ret; + if (!policy) + return -ENODEV; + + cpu_data = policy->driver_data; + cpufreq_cpu_put(policy); ret = cppc_get_desired_perf(cpu, &desired_perf); From 563a6eeecfb813ca532c40ec2403898df20662f9 Mon Sep 17 00:00:00 2001 From: CIQ Kernel Automation Date: Mon, 29 Jun 2026 08:06:03 +0000 Subject: [PATCH 5/9] cpufreq: cppc: Fix invalid return value in .get() callback jira VULN-5484 cve-bf CVE-2024-38573 commit-author Marc Zyngier commit 2b8e6b58889c672e1ae3601d9b2b070be4dc2fbc upstream-diff Also fix hisi_cppc_cpufreq_get_rate() which still exists in this kernel but was removed upstream before this fix was written (removed by ea1829d4d413b). Returning a negative error code in a function with an unsigned return type is a pretty bad idea. It is probably worse when the justification for the change is "our static analisys tool found it". Fixes: cf7de25878a1 ("cppc_cpufreq: Fix possible null pointer dereference") Signed-off-by: Marc Zyngier Cc: "Rafael J. Wysocki" Cc: Viresh Kumar Reviewed-by: Lifeng Zheng Signed-off-by: Viresh Kumar (cherry picked from commit 2b8e6b58889c672e1ae3601d9b2b070be4dc2fbc) Signed-off-by: CIQ Kernel Automation --- drivers/cpufreq/cppc_cpufreq.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/cpufreq/cppc_cpufreq.c b/drivers/cpufreq/cppc_cpufreq.c index 68109bcab89e6..6c3f4e1e45695 100644 --- a/drivers/cpufreq/cppc_cpufreq.c +++ b/drivers/cpufreq/cppc_cpufreq.c @@ -845,7 +845,7 @@ static unsigned int cppc_cpufreq_get_rate(unsigned int cpu) int ret; if (!policy) - return -ENODEV; + return 0; cpu_data = policy->driver_data; @@ -933,7 +933,7 @@ static unsigned int hisi_cppc_cpufreq_get_rate(unsigned int cpu) int ret; if (!policy) - return -ENODEV; + return 0; cpu_data = policy->driver_data; From a008c853818de76df6d6aa542d4ec5255fdf4a4f Mon Sep 17 00:00:00 2001 From: CIQ Kernel Automation Date: Mon, 29 Jun 2026 08:06:50 +0000 Subject: [PATCH 6/9] wifi: nl80211: don't free NULL coalescing rule jira VULN-5407 cve CVE-2024-36941 commit-author Johannes Berg commit 801ea33ae82d6a9d954074fbcf8ea9d18f1543a7 If the parsing fails, we can dereference a NULL pointer here. Cc: stable@vger.kernel.org Fixes: be29b99a9b51 ("cfg80211/nl80211: Add packet coalesce support") Reviewed-by: Miriam Rachel Korenblit Link: https://msgid.link/20240418105220.b328f80406e7.Id75d961050deb05b3e4e354e024866f350c68103@changeid Signed-off-by: Johannes Berg (cherry picked from commit 801ea33ae82d6a9d954074fbcf8ea9d18f1543a7) Signed-off-by: CIQ Kernel Automation --- net/wireless/nl80211.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index 31e0fcd401d7a..486752a19f001 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -13772,6 +13772,8 @@ static int nl80211_set_coalesce(struct sk_buff *skb, struct genl_info *info) error: for (i = 0; i < new_coalesce.n_rules; i++) { tmp_rule = &new_coalesce.rules[i]; + if (!tmp_rule) + continue; for (j = 0; j < tmp_rule->n_patterns; j++) kfree(tmp_rule->patterns[j].mask); kfree(tmp_rule->patterns); From c790c54a017c2e1648301d31c3728c15f8f4e91f Mon Sep 17 00:00:00 2001 From: CIQ Kernel Automation Date: Mon, 29 Jun 2026 08:07:32 +0000 Subject: [PATCH 7/9] icmp: prevent possible NULL dereferences from icmp_build_probe() jira VULN-43056 cve CVE-2024-35857 commit-author Eric Dumazet commit c58e88d49097bd12dfcfef4f075b43f5d5830941 First problem is a double call to __in_dev_get_rcu(), because the second one could return NULL. if (__in_dev_get_rcu(dev) && __in_dev_get_rcu(dev)->ifa_list) Second problem is a read from dev->ip6_ptr with no NULL check: if (!list_empty(&rcu_dereference(dev->ip6_ptr)->addr_list)) Use the correct RCU API to fix these. v2: add missing include Fixes: d329ea5bd884 ("icmp: add response to RFC 8335 PROBE messages") Signed-off-by: Eric Dumazet Cc: Andreas Roeseler Reviewed-by: David Ahern Signed-off-by: David S. Miller (cherry picked from commit c58e88d49097bd12dfcfef4f075b43f5d5830941) Signed-off-by: CIQ Kernel Automation --- net/ipv4/icmp.c | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-) diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c index d47ce05f72583..0e6ab7ddce6b1 100644 --- a/net/ipv4/icmp.c +++ b/net/ipv4/icmp.c @@ -92,6 +92,7 @@ #include #include #include +#include /* * Build xmit assembly blocks @@ -1027,6 +1028,8 @@ bool icmp_build_probe(struct sk_buff *skb, struct icmphdr *icmphdr) struct icmp_ext_hdr *ext_hdr, _ext_hdr; struct icmp_ext_echo_iio *iio, _iio; struct net *net = dev_net(skb->dev); + struct inet6_dev *in6_dev; + struct in_device *in_dev; struct net_device *dev; char buff[IFNAMSIZ]; u16 ident_len; @@ -1110,10 +1113,15 @@ bool icmp_build_probe(struct sk_buff *skb, struct icmphdr *icmphdr) /* Fill bits in reply message */ if (dev->flags & IFF_UP) status |= ICMP_EXT_ECHOREPLY_ACTIVE; - if (__in_dev_get_rcu(dev) && __in_dev_get_rcu(dev)->ifa_list) + + in_dev = __in_dev_get_rcu(dev); + if (in_dev && rcu_access_pointer(in_dev->ifa_list)) status |= ICMP_EXT_ECHOREPLY_IPV4; - if (!list_empty(&rcu_dereference(dev->ip6_ptr)->addr_list)) + + in6_dev = __in6_dev_get(dev); + if (in6_dev && !list_empty(&in6_dev->addr_list)) status |= ICMP_EXT_ECHOREPLY_IPV6; + dev_put(dev); icmphdr->un.echo.sequence |= htons(status); return true; From b51ffa9c2ef6c1bb744a660f7fcdc4f3bd618a3a Mon Sep 17 00:00:00 2001 From: CIQ Kernel Automation Date: Mon, 29 Jun 2026 08:08:17 +0000 Subject: [PATCH 8/9] scsi: qedi: Fix crash while reading debugfs attribute jira VULN-43643 cve CVE-2024-40978 commit-author Manish Rangankar commit 28027ec8e32ecbadcd67623edb290dad61e735b5 The qedi_dbg_do_not_recover_cmd_read() function invokes sprintf() directly on a __user pointer, which results into the crash. To fix this issue, use a small local stack buffer for sprintf() and then call simple_read_from_buffer(), which in turns make the copy_to_user() call. BUG: unable to handle page fault for address: 00007f4801111000 PGD 8000000864df6067 P4D 8000000864df6067 PUD 864df7067 PMD 846028067 PTE 0 Oops: 0002 [#1] PREEMPT SMP PTI Hardware name: HPE ProLiant DL380 Gen10/ProLiant DL380 Gen10, BIOS U30 06/15/2023 RIP: 0010:memcpy_orig+0xcd/0x130 RSP: 0018:ffffb7a18c3ffc40 EFLAGS: 00010202 RAX: 00007f4801111000 RBX: 00007f4801111000 RCX: 000000000000000f RDX: 000000000000000f RSI: ffffffffc0bfd7a0 RDI: 00007f4801111000 RBP: ffffffffc0bfd7a0 R08: 725f746f6e5f6f64 R09: 3d7265766f636572 R10: ffffb7a18c3ffd08 R11: 0000000000000000 R12: 00007f4881110fff R13: 000000007fffffff R14: ffffb7a18c3ffca0 R15: ffffffffc0bfd7af FS: 00007f480118a740(0000) GS:ffff98e38af00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00007f4801111000 CR3: 0000000864b8e001 CR4: 00000000007706e0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 PKRU: 55555554 Call Trace: ? __die_body+0x1a/0x60 ? page_fault_oops+0x183/0x510 ? exc_page_fault+0x69/0x150 ? asm_exc_page_fault+0x22/0x30 ? memcpy_orig+0xcd/0x130 vsnprintf+0x102/0x4c0 sprintf+0x51/0x80 qedi_dbg_do_not_recover_cmd_read+0x2f/0x50 [qedi 6bcfdeeecdea037da47069eca2ba717c84a77324] full_proxy_read+0x50/0x80 vfs_read+0xa5/0x2e0 ? folio_add_new_anon_rmap+0x44/0xa0 ? set_pte_at+0x15/0x30 ? do_pte_missing+0x426/0x7f0 ksys_read+0xa5/0xe0 do_syscall_64+0x58/0x80 ? __count_memcg_events+0x46/0x90 ? count_memcg_event_mm+0x3d/0x60 ? handle_mm_fault+0x196/0x2f0 ? do_user_addr_fault+0x267/0x890 ? exc_page_fault+0x69/0x150 entry_SYSCALL_64_after_hwframe+0x72/0xdc RIP: 0033:0x7f4800f20b4d Tested-by: Martin Hoyer Reviewed-by: John Meneghini Signed-off-by: Manish Rangankar Link: https://lore.kernel.org/r/20240415072155.30840-1-mrangankar@marvell.com Signed-off-by: Martin K. Petersen (cherry picked from commit 28027ec8e32ecbadcd67623edb290dad61e735b5) Signed-off-by: CIQ Kernel Automation --- drivers/scsi/qedi/qedi_debugfs.c | 12 ++++-------- 1 file changed, 4 insertions(+), 8 deletions(-) diff --git a/drivers/scsi/qedi/qedi_debugfs.c b/drivers/scsi/qedi/qedi_debugfs.c index 8deb2001dc2ff..37eed6a278164 100644 --- a/drivers/scsi/qedi/qedi_debugfs.c +++ b/drivers/scsi/qedi/qedi_debugfs.c @@ -120,15 +120,11 @@ static ssize_t qedi_dbg_do_not_recover_cmd_read(struct file *filp, char __user *buffer, size_t count, loff_t *ppos) { - size_t cnt = 0; - - if (*ppos) - return 0; + char buf[64]; + int len; - cnt = sprintf(buffer, "do_not_recover=%d\n", qedi_do_not_recover); - cnt = min_t(int, count, cnt - *ppos); - *ppos += cnt; - return cnt; + len = sprintf(buf, "do_not_recover=%d\n", qedi_do_not_recover); + return simple_read_from_buffer(buffer, count, ppos, buf, len); } static int From a2d6e00e7f4e315d4bf2e5ffa164c22a57905906 Mon Sep 17 00:00:00 2001 From: CIQ Kernel Automation Date: Mon, 29 Jun 2026 08:08:58 +0000 Subject: [PATCH 9/9] mlxsw: spectrum_acl_tcam: Fix memory leak when canceling rehash work jira VULN-43023 cve CVE-2024-35852 commit-author Ido Schimmel commit fb4e2b70a7194b209fc7320bbf33b375f7114bd5 The rehash delayed work is rescheduled with a delay if the number of credits at end of the work is not negative as supposedly it means that the migration ended. Otherwise, it is rescheduled immediately. After "mlxsw: spectrum_acl_tcam: Fix possible use-after-free during rehash" the above is no longer accurate as a non-negative number of credits is no longer indicative of the migration being done. It can also happen if the work encountered an error in which case the migration will resume the next time the work is scheduled. The significance of the above is that it is possible for the work to be pending and associated with hints that were allocated when the migration started. This leads to the hints being leaked [1] when the work is canceled while pending as part of ACL region dismantle. Fix by freeing the hints if hints are associated with a work that was canceled while pending. Blame the original commit since the reliance on not having a pending work associated with hints is fragile. [1] unreferenced object 0xffff88810e7c3000 (size 256): comm "kworker/0:16", pid 176, jiffies 4295460353 hex dump (first 32 bytes): 00 30 95 11 81 88 ff ff 61 00 00 00 00 00 00 80 .0......a....... 00 00 61 00 40 00 00 00 00 00 00 00 04 00 00 00 ..a.@........... backtrace (crc 2544ddb9): [<00000000cf8cfab3>] kmalloc_trace+0x23f/0x2a0 [<000000004d9a1ad9>] objagg_hints_get+0x42/0x390 [<000000000b143cf3>] mlxsw_sp_acl_erp_rehash_hints_get+0xca/0x400 [<0000000059bdb60a>] mlxsw_sp_acl_tcam_vregion_rehash_work+0x868/0x1160 [<00000000e81fd734>] process_one_work+0x59c/0xf20 [<00000000ceee9e81>] worker_thread+0x799/0x12c0 [<00000000bda6fe39>] kthread+0x246/0x300 [<0000000070056d23>] ret_from_fork+0x34/0x70 [<00000000dea2b93e>] ret_from_fork_asm+0x1a/0x30 Fixes: c9c9af91f1d9 ("mlxsw: spectrum_acl: Allow to interrupt/continue rehash work") Signed-off-by: Ido Schimmel Tested-by: Alexander Zubkov Signed-off-by: Petr Machata Reviewed-by: Simon Horman Link: https://lore.kernel.org/r/0cc12ebb07c4d4c41a1265ee2c28b392ff997a86.1713797103.git.petrm@nvidia.com Signed-off-by: Jakub Kicinski (cherry picked from commit fb4e2b70a7194b209fc7320bbf33b375f7114bd5) Signed-off-by: CIQ Kernel Automation --- drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c index 3b9ba8fa247ab..87efdd645d525 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c @@ -879,10 +879,14 @@ mlxsw_sp_acl_tcam_vregion_destroy(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_acl_tcam *tcam = vregion->tcam; if (vgroup->vregion_rehash_enabled && ops->region_rehash_hints_get) { + struct mlxsw_sp_acl_tcam_rehash_ctx *ctx = &vregion->rehash.ctx; + mutex_lock(&tcam->lock); list_del(&vregion->tlist); mutex_unlock(&tcam->lock); - cancel_delayed_work_sync(&vregion->rehash.dw); + if (cancel_delayed_work_sync(&vregion->rehash.dw) && + ctx->hints_priv) + ops->region_rehash_hints_put(ctx->hints_priv); } mlxsw_sp_acl_tcam_vgroup_vregion_detach(mlxsw_sp, vregion); if (vregion->region2)