Linux Kernel

In the Linux kernel, the following vulnerability has been resolved: tracing: Fix memory leak in test_gen_synth_cmd() and test_empty_synth_event() test_gen_synth_cmd() only free buf in fail path, hence buf will leak when there is no failure. Add kfree(buf) to prevent the memleak. The same reason and solution in test_empty_synth_event(). unreferenced object 0xffff8881127de000 (size 2048): comm "modprobe", pid 247, jiffies 4294972316 (age 78.756s) hex dump (first 32 bytes): 20 67 65 6e 5f 73 79 6e 74 68 5f 74 65 73 74 20 gen_synth_test 20 70 69 64 5f 74 20 6e 65 78 74 5f 70 69 64 5f pid_t next_pid_ backtrace: [<000000004254801a>] kmalloc_trace+0x26/0x100 [<0000000039eb1cf5>] 0xffffffffa00083cd [<000000000e8c3bc8>] 0xffffffffa00086ba [<00000000c293d1ea>] do_one_initcall+0xdb/0x480 [<00000000aa189e6d>] do_init_module+0x1cf/0x680 [<00000000d513222b>] load_module+0x6a50/0x70a0 [<000000001fd4d529>] __do_sys_finit_module+0x12f/0x1c0 [<00000000b36c4c0f>] do_syscall_64+0x3f/0x90 [<00000000bbf20cf3>] entry_SYSCALL_64_after_hwframe+0x63/0xcd unreferenced object 0xffff8881127df000 (size 2048): comm "modprobe", pid 247, jiffies 4294972324 (age 78.728s) hex dump (first 32 bytes): 20 65 6d 70 74 79 5f 73 79 6e 74 68 5f 74 65 73 empty_synth_tes 74 20 20 70 69 64 5f 74 20 6e 65 78 74 5f 70 69 t pid_t next_pi backtrace: [<000000004254801a>] kmalloc_trace+0x26/0x100 [<00000000d4db9a3d>] 0xffffffffa0008071 [<00000000c31354a5>] 0xffffffffa00086ce [<00000000c293d1ea>] do_one_initcall+0xdb/0x480 [<00000000aa189e6d>] do_init_module+0x1cf/0x680 [<00000000d513222b>] load_module+0x6a50/0x70a0 [<000000001fd4d529>] __do_sys_finit_module+0x12f/0x1c0 [<00000000b36c4c0f>] do_syscall_64+0x3f/0x90 [<00000000bbf20cf3>] entry_SYSCALL_64_after_hwframe+0x63/0xcdShow less

In the Linux kernel, the following vulnerability has been resolved: tracing: Fix wild-memory-access in register_synth_event() In register_synth_event(), if set_synth_event_print_fmt() failed, then both trace_remove_event_call() and unregister_trace_event() will be called, which means the trace_event_call will call __unregister_trace_event() twice. As the result, the second unregister will causes the wild-memory-access. register_synth_event set_synth_event_print_fmt failed trace_remove_event_call event_remove if call->event.funcs then __unregister_trace_event (first call) unregister_trace_event __unregister_trace_event (second call) Fix the bug by avoiding to call the second __unregister_trace_event() by checking if the first one is called. general protection fault, probably for non-canonical address 0xfbd59c0000000024: 0000 [#1] SMP KASAN PTI KASAN: maybe wild-memory-access in range [0xdead000000000120-0xdead000000000127] CPU: 0 PID: 3807 Comm: modprobe Not tainted 6.1.0-rc1-00186-g76f33a7eedb4 #299 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.15.0-0-g2dd4b9b3f840-prebuilt.qemu.org 04/01/2014 RIP: 0010:unregister_trace_event+0x6e/0x280 Code: 00 fc ff df 4c 89 ea 48 c1 ea 03 80 3c 02 00 0f 85 0e 02 00 00 48 b8 00 00 00 00 00 fc ff df 4c 8b 63 08 4c 89 e2 48 c1 ea 03 <80> 3c 02 00 0f 85 e2 01 00 00 49 89 2c 24 48 85 ed 74 28 e8 7a 9b RSP: 0018:ffff88810413f370 EFLAGS: 00010a06 RAX: dffffc0000000000 RBX: ffff888105d050b0 RCX: 0000000000000000 RDX: 1bd5a00000000024 RSI: ffff888119e276e0 RDI: ffffffff835a8b20 RBP: dead000000000100 R08: 0000000000000000 R09: fffffbfff0913481 R10: ffffffff8489a407 R11: fffffbfff0913480 R12: dead000000000122 R13: ffff888105d050b8 R14: 0000000000000000 R15: ffff888105d05028 FS: 00007f7823e8d540(0000) GS:ffff888119e00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00007f7823e7ebec CR3: 000000010a058002 CR4: 0000000000330ef0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: <TASK> __create_synth_event+0x1e37/0x1eb0 create_or_delete_synth_event+0x110/0x250 synth_event_run_command+0x2f/0x110 test_gen_synth_cmd+0x170/0x2eb [synth_event_gen_test] synth_event_gen_test_init+0x76/0x9bc [synth_event_gen_test] do_one_initcall+0xdb/0x480 do_init_module+0x1cf/0x680 load_module+0x6a50/0x70a0 __do_sys_finit_module+0x12f/0x1c0 do_syscall_64+0x3f/0x90 entry_SYSCALL_64_after_hwframe+0x63/0xcdShow less

In the Linux kernel, the following vulnerability has been resolved: tracing: Fix race where eprobes can be called before the event The flag that tells the event to call its triggers after reading the event is set for eprobes after the eprobe is enabled. This leads to a race where the eprobe may be triggered at the beginning of the event where the record information is NULL. The eprobe then dereferences the NULL record causing a NULL kernel pointer bug. Test for a NULL record to keep this from happening.Show less

In the Linux kernel, the following vulnerability has been resolved: tracing: kprobe: Fix potential null-ptr-deref on trace_event_file in kprobe_event_gen_test_exit() When trace_get_event_file() failed, gen_kretprobe_test will be assigned as the error code. If module kprobe_event_gen_test is removed now, the null pointer dereference will happen in kprobe_event_gen_test_exit(). Check if gen_kprobe_test or gen_kretprobe_test is error code or NULL before dereference them. BUG: kernel NULL pointer dereference, address: 0000000000000012 PGD 0 P4D 0 Oops: 0000 [#1] SMP PTI CPU: 3 PID: 2210 Comm: modprobe Not tainted 6.1.0-rc1-00171-g2159299a3b74-dirty #217 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.15.0-0-g2dd4b9b3f840-prebuilt.qemu.org 04/01/2014 RIP: 0010:kprobe_event_gen_test_exit+0x1c/0xb5 [kprobe_event_gen_test] Code: Unable to access opcode bytes at 0xffffffff9ffffff2. RSP: 0018:ffffc900015bfeb8 EFLAGS: 00010246 RAX: ffffffffffffffea RBX: ffffffffa0002080 RCX: 0000000000000000 RDX: ffffffffa0001054 RSI: ffffffffa0001064 RDI: ffffffffdfc6349c RBP: ffffffffa0000000 R08: 0000000000000004 R09: 00000000001e95c0 R10: 0000000000000000 R11: 0000000000000001 R12: 0000000000000800 R13: ffffffffa0002420 R14: 0000000000000000 R15: 0000000000000000 FS: 00007f56b75be540(0000) GS:ffff88813bc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: ffffffff9ffffff2 CR3: 000000010874a006 CR4: 0000000000330ee0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: <TASK> __x64_sys_delete_module+0x206/0x380 ? lockdep_hardirqs_on_prepare+0xd8/0x190 ? syscall_enter_from_user_mode+0x1c/0x50 do_syscall_64+0x3f/0x90 entry_SYSCALL_64_after_hwframe+0x63/0xcdShow less

In the Linux kernel, the following vulnerability has been resolved: tracing: kprobe: Fix potential null-ptr-deref on trace_array in kprobe_event_gen_test_exit() When test_gen_kprobe_cmd() failed after kprobe_event_gen_cmd_end(), it will goto delete, which will call kprobe_event_delete() and release the corresponding resource. However, the trace_array in gen_kretprobe_test will point to the invalid resource. Set gen_kretprobe_test to NULL after called kprobe_event_delete() to prevent null-ptr-deref. BUG: kernel NULL pointer dereference, address: 0000000000000070 PGD 0 P4D 0 Oops: 0000 [#1] SMP PTI CPU: 0 PID: 246 Comm: modprobe Tainted: G W 6.1.0-rc1-00174-g9522dc5c87da-dirty #248 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.15.0-0-g2dd4b9b3f840-prebuilt.qemu.org 04/01/2014 RIP: 0010:__ftrace_set_clr_event_nolock+0x53/0x1b0 Code: e8 82 26 fc ff 49 8b 1e c7 44 24 0c ea ff ff ff 49 39 de 0f 84 3c 01 00 00 c7 44 24 18 00 00 00 00 e8 61 26 fc ff 48 8b 6b 10 <44> 8b 65 70 4c 8b 6d 18 41 f7 c4 00 02 00 00 75 2f RSP: 0018:ffffc9000159fe00 EFLAGS: 00010293 RAX: 0000000000000000 RBX: ffff88810971d268 RCX: 0000000000000000 RDX: ffff8881080be600 RSI: ffffffff811b48ff RDI: ffff88810971d058 RBP: 0000000000000000 R08: 0000000000000000 R09: 0000000000000001 R10: ffffc9000159fe58 R11: 0000000000000001 R12: ffffffffa0001064 R13: ffffffffa000106c R14: ffff88810971d238 R15: 0000000000000000 FS: 00007f89eeff6540(0000) GS:ffff88813b600000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000000000000070 CR3: 000000010599e004 CR4: 0000000000330ef0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: <TASK> __ftrace_set_clr_event+0x3e/0x60 trace_array_set_clr_event+0x35/0x50 ? 0xffffffffa0000000 kprobe_event_gen_test_exit+0xcd/0x10b [kprobe_event_gen_test] __x64_sys_delete_module+0x206/0x380 ? lockdep_hardirqs_on_prepare+0xd8/0x190 ? syscall_enter_from_user_mode+0x1c/0x50 do_syscall_64+0x3f/0x90 entry_SYSCALL_64_after_hwframe+0x63/0xcd RIP: 0033:0x7f89eeb061b7Show less

In the Linux kernel, the following vulnerability has been resolved: rethook: fix a potential memleak in rethook_alloc() In rethook_alloc(), the variable rh is not freed or passed out if handler is NULL, which could lead to a memleak, fix it. [Masami: Add "rethook:" tag to the title.] Acke-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>Show less

In the Linux kernel, the following vulnerability has been resolved: iio: adc: at91_adc: fix possible memory leak in at91_adc_allocate_trigger() If iio_trigger_register() returns error, it should call iio_trigger_free() to give up the reference that hold in iio_trigger_alloc(), so that it can call iio_trig_release() to free memory when the refcount hit to 0.Show less

In the Linux kernel, the following vulnerability has been resolved: iio: trigger: sysfs: fix possible memory leak in iio_sysfs_trig_init() dev_set_name() allocates memory for name, it need be freed when device_add() fails, call put_device() to give up the reference that hold in device_initialize(), so that it can be freed in kobject_cleanup() when the refcount hit to 0. Fault injection test can trigger this: unreferenced object 0xffff8e8340a7b4c0 (size 32): comm "modprobe", pid 243, jiffies 4294678145 (age 48.845s) hex dump (first 32 bytes): 69 69 6f 5f 73 79 73 66 73 5f 74 72 69 67 67 65 iio_sysfs_trigge 72 00 a7 40 83 8e ff ff 00 86 13 c4 f6 ee ff ff r..@............ backtrace: [<0000000074999de8>] __kmem_cache_alloc_node+0x1e9/0x360 [<00000000497fd30b>] __kmalloc_node_track_caller+0x44/0x1a0 [<000000003636c520>] kstrdup+0x2d/0x60 [<0000000032f84da2>] kobject_set_name_vargs+0x1e/0x90 [<0000000092efe493>] dev_set_name+0x4e/0x70Show less

In the Linux kernel, the following vulnerability has been resolved: io_uring: fix multishot accept request leaks Having REQ_F_POLLED set doesn't guarantee that the request is executed as a multishot from the polling path. Fortunately for us, if the code thinks it's multishot issue when it's not, it can only ask to skip completion so leaking the request. Use issue_flags to mark multipoll issues.Show less

In the Linux kernel, the following vulnerability has been resolved: Input: iforce - invert valid length check when fetching device IDs syzbot is reporting uninitialized value at iforce_init_device() [1], for commit 6ac0aec6b0a6 ("Input: iforce - allow callers supply data buffer when fetching device IDs") is checking that valid length is shorter than bytes to read. Since iforce_get_id_packet() stores valid length when returning 0, the caller needs to check that valid length is longer than or equals to bytes to read.Show less

In the Linux kernel, the following vulnerability has been resolved: scsi: zfcp: Fix double free of FSF request when qdio send fails We used to use the wrong type of integer in 'zfcp_fsf_req_send()' to cache the FSF request ID when sending a new FSF request. This is used in case the sending fails and we need to remove the request from our internal hash table again (so we don't keep an invalid reference and use it when we free the request again). In 'zfcp_fsf_req_send()' we used to cache the ID as 'int' (signed and 32 bit wide), but the rest of the zfcp code (and the firmware specification) handles the ID as 'unsigned long'/'u64' (unsigned and 64 bit wide [s390x ELF ABI]). For one this has the obvious problem that when the ID grows past 32 bit (this can happen reasonably fast) it is truncated to 32 bit when storing it in the cache variable and so doesn't match the original ID anymore. The second less obvious problem is that even when the original ID has not yet grown past 32 bit, as soon as the 32nd bit is set in the original ID (0x80000000 = 2'147'483'648) we will have a mismatch when we cast it back to 'unsigned long'. As the cached variable is of a signed type, the compiler will choose a sign-extending instruction to load the 32 bit variable into a 64 bit register (e.g.: 'lgf %r11,188(%r15)'). So once we pass the cached variable into 'zfcp_reqlist_find_rm()' to remove the request again all the leading zeros will be flipped to ones to extend the sign and won't match the original ID anymore (this has been observed in practice). If we can't successfully remove the request from the hash table again after 'zfcp_qdio_send()' fails (this happens regularly when zfcp cannot notify the adapter about new work because the adapter is already gone during e.g. a ChpID toggle) we will end up with a double free. We unconditionally free the request in the calling function when 'zfcp_fsf_req_send()' fails, but because the request is still in the hash table we end up with a stale memory reference, and once the zfcp adapter is either reset during recovery or shutdown we end up freeing the same memory twice. The resulting stack traces vary depending on the kernel and have no direct correlation to the place where the bug occurs. Here are three examples that have been seen in practice: list_del corruption. next->prev should be 00000001b9d13800, but was 00000000dead4ead. (next=00000001bd131a00) ------------[ cut here ]------------ kernel BUG at lib/list_debug.c:62! monitor event: 0040 ilc:2 [#1] PREEMPT SMP Modules linked in: ... CPU: 9 PID: 1617 Comm: zfcperp0.0.1740 Kdump: loaded Hardware name: ... Krnl PSW : 0704d00180000000 00000003cbeea1f8 (__list_del_entry_valid+0x98/0x140) R:0 T:1 IO:1 EX:1 Key:0 M:1 W:0 P:0 AS:3 CC:1 PM:0 RI:0 EA:3 Krnl GPRS: 00000000916d12f1 0000000080000000 000000000000006d 00000003cb665cd6 0000000000000001 0000000000000000 0000000000000000 00000000d28d21e8 00000000d3844000 00000380099efd28 00000001bd131a00 00000001b9d13800 00000000d3290100 0000000000000000 00000003cbeea1f4 00000380099efc70 Krnl Code: 00000003cbeea1e8: c020004f68a7 larl %r2,00000003cc8d7336 00000003cbeea1ee: c0e50027fd65 brasl %r14,00000003cc3e9cb8 #00000003cbeea1f4: af000000 mc 0,0 >00000003cbeea1f8: c02000920440 larl %r2,00000003cd12aa78 00000003cbeea1fe: c0e500289c25 brasl %r14,00000003cc3fda48 00000003cbeea204: b9040043 lgr %r4,%r3 00000003cbeea208: b9040051 lgr %r5,%r1 00000003cbeea20c: b9040032 lgr %r3,%r2 Call Trace: [<00000003cbeea1f8>] __list_del_entry_valid+0x98/0x140 ([<00000003cbeea1f4>] __list_del_entry_valid+0x94/0x140) [<000003ff7ff502fe>] zfcp_fsf_req_dismiss_all+0xde/0x150 [zfcp] [<000003ff7ff49cd0>] zfcp_erp_strategy_do_action+0x160/0x280 [zfcp] ---truncated---Show less

In the Linux kernel, the following vulnerability has been resolved: misc/vmw_vmci: fix an infoleak in vmci_host_do_receive_datagram() `struct vmci_event_qp` allocated by qp_notify_peer() contains padding, which may carry uninitialized data to the userspace, as observed by KMSAN: BUG: KMSAN: kernel-infoleak in instrument_copy_to_user ./include/linux/instrumented.h:121 instrument_copy_to_user ./include/linux/instrumented.h:121 _copy_to_user+0x5f/0xb0 lib/usercopy.c:33 copy_to_user ./include/linux/uaccess.h:169 vmci_host_do_receive_datagram drivers/misc/vmw_vmci/vmci_host.c:431 vmci_host_unlocked_ioctl+0x33d/0x43d0 drivers/misc/vmw_vmci/vmci_host.c:925 vfs_ioctl fs/ioctl.c:51 ... Uninit was stored to memory at: kmemdup+0x74/0xb0 mm/util.c:131 dg_dispatch_as_host drivers/misc/vmw_vmci/vmci_datagram.c:271 vmci_datagram_dispatch+0x4f8/0xfc0 drivers/misc/vmw_vmci/vmci_datagram.c:339 qp_notify_peer+0x19a/0x290 drivers/misc/vmw_vmci/vmci_queue_pair.c:1479 qp_broker_attach drivers/misc/vmw_vmci/vmci_queue_pair.c:1662 qp_broker_alloc+0x2977/0x2f30 drivers/misc/vmw_vmci/vmci_queue_pair.c:1750 vmci_qp_broker_alloc+0x96/0xd0 drivers/misc/vmw_vmci/vmci_queue_pair.c:1940 vmci_host_do_alloc_queuepair drivers/misc/vmw_vmci/vmci_host.c:488 vmci_host_unlocked_ioctl+0x24fd/0x43d0 drivers/misc/vmw_vmci/vmci_host.c:927 ... Local variable ev created at: qp_notify_peer+0x54/0x290 drivers/misc/vmw_vmci/vmci_queue_pair.c:1456 qp_broker_attach drivers/misc/vmw_vmci/vmci_queue_pair.c:1662 qp_broker_alloc+0x2977/0x2f30 drivers/misc/vmw_vmci/vmci_queue_pair.c:1750 Bytes 28-31 of 48 are uninitialized Memory access of size 48 starts at ffff888035155e00 Data copied to user address 0000000020000100 Use memset() to prevent the infoleaks. Also speculatively fix qp_notify_peer_local(), which may suffer from the same problem.Show less

In the Linux kernel, the following vulnerability has been resolved: mmc: sdhci-pci: Fix possible memory leak caused by missing pci_dev_put() pci_get_device() will increase the reference count for the returned pci_dev. We need to use pci_dev_put() to decrease the reference count before amd_probe() returns. There is no problem for the 'smbus_dev == NULL' branch because pci_dev_put() can also handle the NULL input parameter case.Show less

In the Linux kernel, the following vulnerability has been resolved: blk-cgroup: properly pin the parent in blkcg_css_online blkcg_css_online is supposed to pin the blkcg of the parent, but 397c9f46ee4d refactored things and along the way, changed it to pin the css instead. This results in extra pins, and we end up leaking blkcgs and cgroups.Show less

In the Linux kernel, the following vulnerability has been resolved: x86/sgx: Add overflow check in sgx_validate_offset_length() sgx_validate_offset_length() function verifies "offset" and "length" arguments provided by userspace, but was missing an overflow check on their addition. Add it.Show less

In the Linux kernel, the following vulnerability has been resolved: perf/x86/amd/uncore: Fix memory leak for events array When a CPU comes online, the per-CPU NB and LLC uncore contexts are freed but not the events array within the context structure. This causes a memory leak as identified by the kmemleak detector. [...] unreferenced object 0xffff8c5944b8e320 (size 32): comm "swapper/0", pid 1, jiffies 4294670387 (age 151.072s) hex dump (first 32 bytes): 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................ 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................ backtrace: [<000000000759fb79>] amd_uncore_cpu_up_prepare+0xaf/0x230 [<00000000ddc9e126>] cpuhp_invoke_callback+0x2cf/0x470 [<0000000093e727d4>] cpuhp_issue_call+0x14d/0x170 [<0000000045464d54>] __cpuhp_setup_state_cpuslocked+0x11e/0x330 [<0000000069f67cbd>] __cpuhp_setup_state+0x6b/0x110 [<0000000015365e0f>] amd_uncore_init+0x260/0x321 [<00000000089152d2>] do_one_initcall+0x3f/0x1f0 [<000000002d0bd18d>] kernel_init_freeable+0x1ca/0x212 [<0000000030be8dde>] kernel_init+0x11/0x120 [<0000000059709e59>] ret_from_fork+0x22/0x30 unreferenced object 0xffff8c5944b8dd40 (size 64): comm "swapper/0", pid 1, jiffies 4294670387 (age 151.072s) hex dump (first 32 bytes): 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................ 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................ backtrace: [<00000000306efe8b>] amd_uncore_cpu_up_prepare+0x183/0x230 [<00000000ddc9e126>] cpuhp_invoke_callback+0x2cf/0x470 [<0000000093e727d4>] cpuhp_issue_call+0x14d/0x170 [<0000000045464d54>] __cpuhp_setup_state_cpuslocked+0x11e/0x330 [<0000000069f67cbd>] __cpuhp_setup_state+0x6b/0x110 [<0000000015365e0f>] amd_uncore_init+0x260/0x321 [<00000000089152d2>] do_one_initcall+0x3f/0x1f0 [<000000002d0bd18d>] kernel_init_freeable+0x1ca/0x212 [<0000000030be8dde>] kernel_init+0x11/0x120 [<0000000059709e59>] ret_from_fork+0x22/0x30 [...] Fix the problem by freeing the events array before freeing the uncore context.Show less

In the Linux kernel, the following vulnerability has been resolved: x86/fpu: Drop fpregs lock before inheriting FPU permissions Mike Galbraith reported the following against an old fork of preempt-rt but the same issue also applies to the current preempt-rt tree. BUG: sleeping function called from invalid context at kernel/locking/spinlock_rt.c:46 in_atomic(): 1, irqs_disabled(): 0, non_block: 0, pid: 1, name: systemd preempt_count: 1, expected: 0 RCU nest depth: 0, expected: 0 Preemption disabled at: fpu_clone CPU: 6 PID: 1 Comm: systemd Tainted: G E (unreleased) Call Trace: <TASK> dump_stack_lvl ? fpu_clone __might_resched rt_spin_lock fpu_clone ? copy_thread ? copy_process ? shmem_alloc_inode ? kmem_cache_alloc ? kernel_clone ? __do_sys_clone ? do_syscall_64 ? __x64_sys_rt_sigprocmask ? syscall_exit_to_user_mode ? do_syscall_64 ? syscall_exit_to_user_mode ? do_syscall_64 ? syscall_exit_to_user_mode ? do_syscall_64 ? exc_page_fault ? entry_SYSCALL_64_after_hwframe </TASK> Mike says: The splat comes from fpu_inherit_perms() being called under fpregs_lock(), and us reaching the spin_lock_irq() therein due to fpu_state_size_dynamic() returning true despite static key __fpu_state_size_dynamic having never been enabled. Mike's assessment looks correct. fpregs_lock on a PREEMPT_RT kernel disables preemption so calling spin_lock_irq() in fpu_inherit_perms() is unsafe. This problem exists since commit 9e798e9aa14c ("x86/fpu: Prepare fpu_clone() for dynamically enabled features"). Even though the original bug report should not have enabled the paths at all, the bug still exists. fpregs_lock is necessary when editing the FPU registers or a task's FP state but it is not necessary for fpu_inherit_perms(). The only write of any FP state in fpu_inherit_perms() is for the new child which is not running yet and cannot context switch or be borrowed by a kernel thread yet. Hence, fpregs_lock is not protecting anything in the new child until clone() completes and can be dropped earlier. The siglock still needs to be acquired by fpu_inherit_perms() as the read of the parent's permissions has to be serialised. [ bp: Cleanup splat. ]Show less

In the Linux kernel, the following vulnerability has been resolved: perf: Improve missing SIGTRAP checking To catch missing SIGTRAP we employ a WARN in __perf_event_overflow(), which fires if pending_sigtrap was already set: returning to user space without consuming pending_sigtrap, and then having the event fire again would re-enter the kernel and trigger the WARN. This, however, seemed to miss the case where some events not associated with progress in the user space task can fire and the interrupt handler runs before the IRQ work meant to consume pending_sigtrap (and generate the SIGTRAP). syzbot gifted us this stack trace: | WARNING: CPU: 0 PID: 3607 at kernel/events/core.c:9313 __perf_event_overflow | Modules linked in: | CPU: 0 PID: 3607 Comm: syz-executor100 Not tainted 6.1.0-rc2-syzkaller-00073-g88619e77b33d #0 | Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 10/11/2022 | RIP: 0010:__perf_event_overflow+0x498/0x540 kernel/events/core.c:9313 | <...> | Call Trace: | <TASK> | perf_swevent_hrtimer+0x34f/0x3c0 kernel/events/core.c:10729 | __run_hrtimer kernel/time/hrtimer.c:1685 [inline] | __hrtimer_run_queues+0x1c6/0xfb0 kernel/time/hrtimer.c:1749 | hrtimer_interrupt+0x31c/0x790 kernel/time/hrtimer.c:1811 | local_apic_timer_interrupt arch/x86/kernel/apic/apic.c:1096 [inline] | __sysvec_apic_timer_interrupt+0x17c/0x640 arch/x86/kernel/apic/apic.c:1113 | sysvec_apic_timer_interrupt+0x40/0xc0 arch/x86/kernel/apic/apic.c:1107 | asm_sysvec_apic_timer_interrupt+0x16/0x20 arch/x86/include/asm/idtentry.h:649 | <...> | </TASK> In this case, syzbot produced a program with event type PERF_TYPE_SOFTWARE and config PERF_COUNT_SW_CPU_CLOCK. The hrtimer manages to fire again before the IRQ work got a chance to run, all while never having returned to user space. Improve the WARN to check for real progress in user space: approximate this by storing a 32-bit hash of the current IP into pending_sigtrap, and if an event fires while pending_sigtrap still matches the previous IP, we assume no progress (false negatives are possible given we could return to user space and trigger again on the same IP).Show less

In the Linux kernel, the following vulnerability has been resolved: perf/x86/amd: Fix crash due to race between amd_pmu_enable_all, perf NMI and throttling amd_pmu_enable_all() does: if (!test_bit(idx, cpuc->active_mask)) continue; amd_pmu_enable_event(cpuc->events[idx]); A perf NMI of another event can come between these two steps. Perf NMI handler internally disables and enables _all_ events, including the one which nmi-intercepted amd_pmu_enable_all() was in process of enabling. If that unintentionally enabled event has very low sampling period and causes immediate successive NMI, causing the event to be throttled, cpuc->events[idx] and cpuc->active_mask gets cleared by x86_pmu_stop(). This will result in amd_pmu_enable_event() getting called with event=NULL when amd_pmu_enable_all() resumes after handling the NMIs. This causes a kernel crash: BUG: kernel NULL pointer dereference, address: 0000000000000198 #PF: supervisor read access in kernel mode #PF: error_code(0x0000) - not-present page [...] Call Trace: <TASK> amd_pmu_enable_all+0x68/0xb0 ctx_resched+0xd9/0x150 event_function+0xb8/0x130 ? hrtimer_start_range_ns+0x141/0x4a0 ? perf_duration_warn+0x30/0x30 remote_function+0x4d/0x60 __flush_smp_call_function_queue+0xc4/0x500 flush_smp_call_function_queue+0x11d/0x1b0 do_idle+0x18f/0x2d0 cpu_startup_entry+0x19/0x20 start_secondary+0x121/0x160 secondary_startup_64_no_verify+0xe5/0xeb </TASK> amd_pmu_disable_all()/amd_pmu_enable_all() calls inside perf NMI handler were recently added as part of BRS enablement but I'm not sure whether we really need them. We can just disable BRS in the beginning and enable it back while returning from NMI. This will solve the issue by not enabling those events whose active_masks are set but are not yet enabled in hw pmu.Show less