kernel.git/arch/parisc/kernel/unwind.c, branch linux-6.19.y

parisc: Avoid crash due to unaligned access in unwinder

2025-11-04T11:21:59Z

Guenter Roeck reported this kernel crash on his emulated B160L machine: Starting network: udhcpc: started, v1.36.1 Backtrace: [<104320d4>] unwind_once+0x1c/0x5c [<10434a00>] walk_stackframe.isra.0+0x74/0xb8 [<10434a6c>] arch_stack_walk+0x28/0x38 [<104e5efc>] stack_trace_save+0x48/0x5c [<105d1bdc>] set_track_prepare+0x44/0x6c [<105d9c80>] ___slab_alloc+0xfc4/0x1024 [<105d9d38>] __slab_alloc.isra.0+0x58/0x90 [<105dc80c>] kmem_cache_alloc_noprof+0x2ac/0x4a0 [<105b8e54>] __anon_vma_prepare+0x60/0x280 [<105a823c>] __vmf_anon_prepare+0x68/0x94 [<105a8b34>] do_wp_page+0x8cc/0xf10 [<105aad88>] handle_mm_fault+0x6c0/0xf08 [<10425568>] do_page_fault+0x110/0x440 [<10427938>] handle_interruption+0x184/0x748 [<11178398>] schedule+0x4c/0x190 BUG: spinlock recursion on CPU#0, ifconfig/2420 lock: terminate_lock.2+0x0/0x1c, .magic: dead4ead, .owner: ifconfig/2420, .owner_cpu: 0 While creating the stack trace, the unwinder uses the stack pointer to guess the previous frame to read the previous stack pointer from memory. The crash happens, because the unwinder tries to read from unaligned memory and as such triggers the unalignment trap handler which then leads to the spinlock recursion and finally to a deadlock. Fix it by checking the alignment before accessing the memory. Reported-by: Guenter Roeck Signed-off-by: Helge Deller Tested-by: Guenter Roeck Cc: stable@vger.kernel.org # v6.12+

parisc: Fix stack unwinder

2024-02-19T20:55:22Z

Debugging shows a large number of unaligned access traps in the unwinder code. Code analysis reveals a number of issues with this code: - handle_interruption is passed twice through dereference_kernel_function_descriptor() - ret_from_kernel_thread, syscall_exit, intr_return, _switch_to_ret, and _call_on_stack are passed through dereference_kernel_function_descriptor() even though they are not declared as function pointers. To fix the problems, drop one of the calls to dereference_kernel_function_descriptor() for handle_interruption, and compare the other pointers directly. Fixes: 6414b30b39f9 ("parisc: unwind: Avoid missing prototype warning for handle_interruption()") Fixes: 8e0ba125c2bf ("parisc/unwind: fix unwinder when CONFIG_64BIT is enabled") Cc: Helge Deller Cc: Sven Schnelle Cc: John David Anglin Cc: Charlie Jenkins Cc: David Laight Signed-off-by: Guenter Roeck Signed-off-by: Helge Deller

parisc: unwind: Avoid missing prototype warning for handle_interruption()

2023-07-03T16:56:03Z

Signed-off-by: Helge Deller

parisc: unwind: Mark start and stop variables __maybe_unused

2023-06-30T15:14:15Z

Signed-off-by: Helge Deller

parisc: Move thread_info into task struct

2021-11-01T06:35:59Z

This implements the CONFIG_THREAD_INFO_IN_TASK option. With this change: - before thread_info was part of the stack and located at the beginning of the stack - now the thread_info struct is moved and located inside the task_struct structure - the stack is allocated and handled like the major other platforms - drop the cpu field of thread_info and use instead the one in task_struct Signed-off-by: Helge Deller Signed-off-by: Sven Schnelle

parisc/unwind: fix unwinder when CONFIG_64BIT is enabled

2021-11-01T06:35:58Z

With 64 bit kernels unwind_special() is not working because it compares the pc to the address of the function descriptor. Add a helper function that compares pc with the dereferenced address. This fixes all of the backtraces on my c8000. Without this changes, a lot of backtraces are missing in kdb or the show-all-tasks command from /proc/sysrq-trigger. Signed-off-by: Sven Schnelle Signed-off-by: Helge Deller

parisc/unwind: use copy_from_kernel_nofault()

2021-10-30T21:11:00Z

I have no idea why get_user() is used there, but we're unwinding the kernel stack, so we should use copy_from_kernel_nofault(). Signed-off-by: Sven Schnelle Signed-off-by: Helge Deller

parisc: Use __ro_after_init in unwind.c

2019-05-10T19:00:45Z

Signed-off-by: Helge Deller

parisc: Fix uninitialized variable usage in unwind.c

2018-10-16T09:37:29Z

As noticed by Dave Anglin, the last commit introduced a small bug where the potentially uninitialized r struct is used instead of the regs pointer as input for unwind_frame_init(). Fix it. Signed-off-by: Helge Deller Reported-by: John David Anglin

parisc: Fix boot failure of 64-bit kernel

2018-08-21T12:32:44Z

Commit c8921d72e390 ("parisc: Fix and improve kernel stack unwinding") broke booting of 64-bit kernels. On 64-bit kernels function pointers are actually function descriptors which require dereferencing. In this patch we instead declare functions in assembly code which are referenced from C-code as external data pointers with the ENTRY() macro and thus can use a simple external reference to the functions. Signed-off-by: Helge Deller Fixes: c8921d72e390 ("parisc: Fix and improve kernel stack unwinding")