kernel.git/arch/riscv/kernel/mcount.S, branch linux-6.19.y

riscv: fgraph: Fix stack layout to match __arch_ftrace_regs argument of ftrace_return_to_handler

2025-03-19T12:03:25Z

Naresh Kamboju reported a "Bad frame pointer" kernel warning while running LTP trace ftrace_stress_test.sh in riscv. We can reproduce the same issue with the following command: ``` $ cd /sys/kernel/debug/tracing $ echo 'f:myprobe do_nanosleep%return args1=$retval' > dynamic_events $ echo 1 > events/fprobes/enable $ echo 1 > tracing_on $ sleep 1 ``` And we can get the following kernel warning: [ 127.692888] ------------[ cut here ]------------ [ 127.693755] Bad frame pointer: expected ff2000000065be50, received ba34c141e9594000 [ 127.693755] from func do_nanosleep return to ffffffff800ccb16 [ 127.698699] WARNING: CPU: 1 PID: 129 at kernel/trace/fgraph.c:755 ftrace_return_to_handler+0x1b2/0x1be [ 127.699894] Modules linked in: [ 127.700908] CPU: 1 UID: 0 PID: 129 Comm: sleep Not tainted 6.14.0-rc3-g0ab191c74642 #32 [ 127.701453] Hardware name: riscv-virtio,qemu (DT) [ 127.701859] epc : ftrace_return_to_handler+0x1b2/0x1be [ 127.702032] ra : ftrace_return_to_handler+0x1b2/0x1be [ 127.702151] epc : ffffffff8013b5e0 ra : ffffffff8013b5e0 sp : ff2000000065bd10 [ 127.702221] gp : ffffffff819c12f8 tp : ff60000080853100 t0 : 6e00000000000000 [ 127.702284] t1 : 0000000000000020 t2 : 6e7566206d6f7266 s0 : ff2000000065bd80 [ 127.702346] s1 : ff60000081262000 a0 : 000000000000007b a1 : ffffffff81894f20 [ 127.702408] a2 : 0000000000000010 a3 : fffffffffffffffe a4 : 0000000000000000 [ 127.702470] a5 : 0000000000000000 a6 : 0000000000000008 a7 : 0000000000000038 [ 127.702530] s2 : ba34c141e9594000 s3 : 0000000000000000 s4 : ff2000000065bdd0 [ 127.702591] s5 : 00007fff8adcf400 s6 : 000055556dc1d8c0 s7 : 0000000000000068 [ 127.702651] s8 : 00007fff8adf5d10 s9 : 000000000000006d s10: 0000000000000001 [ 127.702710] s11: 00005555737377c8 t3 : ffffffff819d899e t4 : ffffffff819d899e [ 127.702769] t5 : ffffffff819d89a0 t6 : ff2000000065bb18 [ 127.702826] status: 0000000200000120 badaddr: 0000000000000000 cause: 0000000000000003 [ 127.703292] [] ftrace_return_to_handler+0x1b2/0x1be [ 127.703760] [] return_to_handler+0x16/0x26 [ 127.704009] [] return_to_handler+0x0/0x26 [ 127.704057] [] common_nsleep+0x42/0x54 [ 127.704117] [] __riscv_sys_clock_nanosleep+0xba/0x10a [ 127.704176] [] do_trap_ecall_u+0x188/0x218 [ 127.704295] [] handle_exception+0x14a/0x156 [ 127.705436] ---[ end trace 0000000000000000 ]--- The reason is that the stack layout for constructing argument for the ftrace_return_to_handler in the return_to_handler does not match the __arch_ftrace_regs structure of riscv, leading to unexpected results. Fixes: a3ed4157b7d8 ("fgraph: Replace fgraph_ret_regs with ftrace_regs") Reported-by: Linux Kernel Functional Testing Closes: https://lore.kernel.org/all/CA+G9fYvp_oAxeDFj88Tk2rfEZ7jtYKAKSwfYS66=57Db9TBdyA@mail.gmail.com Signed-off-by: Pu Lehui Reviewed-by: Alexandre Ghiti Tested-by: Björn Töpel Reviewed-by: Masami Hiramatsu (Google) Link: https://lore.kernel.org/r/20250317031214.4138436-2-pulehui@huaweicloud.com Signed-off-by: Alexandre Ghiti

fgraph: Replace fgraph_ret_regs with ftrace_regs

2024-12-26T15:50:02Z

Use ftrace_regs instead of fgraph_ret_regs for tracing return value on function_graph tracer because of simplifying the callback interface. The CONFIG_HAVE_FUNCTION_GRAPH_RETVAL is also replaced by CONFIG_HAVE_FUNCTION_GRAPH_FREGS. Signed-off-by: Masami Hiramatsu (Google) Acked-by: Heiko Carstens Acked-by: Will Deacon Cc: Catalin Marinas Cc: Alexei Starovoitov Cc: Florent Revest Cc: Martin KaFai Lau Cc: bpf Cc: Alexei Starovoitov Cc: Jiri Olsa Cc: Alan Maguire Cc: Mark Rutland Cc: Huacai Chen Cc: WANG Xuerui Cc: Paul Walmsley Cc: Palmer Dabbelt Cc: Albert Ou Cc: Vasily Gorbik Cc: Alexander Gordeev Cc: Heiko Carstens Cc: Christian Borntraeger Cc: Sven Schnelle Cc: Thomas Gleixner Cc: Ingo Molnar Cc: Borislav Petkov Cc: Dave Hansen Cc: x86@kernel.org Cc: "H. Peter Anvin" Cc: Mathieu Desnoyers Link: https://lore.kernel.org/173518991508.391279.16635322774382197642.stgit@devnote2 Signed-off-by: Steven Rostedt (Google)

riscv: remove MCOUNT_NAME workaround

2024-02-22T23:38:54Z

Now that the minimum supported version of LLVM for building the kernel has been bumped to 13.0.1, the condition for using _mcount as MCOUNT_NAME is always true, as the build will fail during the configuration stage for older LLVM versions. Replace MCOUNT_NAME with _mcount directly. This effectively reverts commit 7ce047715030 ("riscv: Workaround mcount name prior to clang-13"). Link: https://lkml.kernel.org/r/20240125-bump-min-llvm-ver-to-13-0-1-v1-7-f5ff9bda41c5@kernel.org Signed-off-by: Nathan Chancellor Reviewed-by: Kees Cook Cc: Albert Ou Cc: "Aneesh Kumar K.V (IBM)" Cc: Ard Biesheuvel Cc: Borislav Petkov (AMD) Cc: Catalin Marinas Cc: Conor Dooley Cc: Dave Hansen Cc: Ingo Molnar Cc: Mark Rutland Cc: Masahiro Yamada Cc: Michael Ellerman Cc: "Naveen N. Rao" Cc: Nicholas Piggin Cc: Nicolas Schier Cc: Palmer Dabbelt Cc: Paul Walmsley Cc: Russell King Cc: Thomas Gleixner Cc: Will Deacon Signed-off-by: Andrew Morton

use linux/export.h rather than asm-generic/export.h

2024-01-10T04:10:40Z

asm-generic/export.h is a wrapper for linux/export.h, with explicit request to use linux/export.h directly. Signed-off-by: Al Viro Link: https://lore.kernel.org/r/20231214191922.GQ1674809@ZenIV Signed-off-by: Palmer Dabbelt

riscv: Use SYM_*() assembly macros instead of deprecated ones

2023-11-06T17:42:47Z

ENTRY()/END()/WEAK() macros are deprecated and we should make use of the new SYM_*() macros [1] for better annotation of symbols. Replace the deprecated ones with the new ones and fix wrong usage of END()/ENDPROC() to correctly describe the symbols. [1] https://docs.kernel.org/core-api/asm-annotations.html Signed-off-by: Clément Léger Reviewed-by: Andrew Jones Link: https://lore.kernel.org/r/20231024132655.730417-3-cleger@rivosinc.com Signed-off-by: Palmer Dabbelt

riscv: use ".L" local labels in assembly when applicable

2023-11-06T17:42:05Z

For the sake of coherency, use local labels in assembly when applicable. This also avoid kprobes being confused when applying a kprobe since the size of function is computed by checking where the next visible symbol is located. This might end up in computing some function size to be way shorter than expected and thus failing to apply kprobes to the specified offset. Signed-off-by: Clément Léger Reviewed-by: Andrew Jones Link: https://lore.kernel.org/r/20231024132655.730417-2-cleger@rivosinc.com Signed-off-by: Palmer Dabbelt

riscv: Add ftrace_stub_graph

2023-08-23T21:16:38Z

Commit 883bbbffa5a4 ("ftrace,kcfi: Separate ftrace_stub() and ftrace_stub_graph()") added a separate ftrace_stub_graph function for CFI_CLANG. Add the stub to fix FUNCTION_GRAPH_TRACER compatibility with CFI. Reviewed-by: Kees Cook Tested-by: Nathan Chancellor Signed-off-by: Sami Tolvanen Link: https://lore.kernel.org/r/20230710183544.999540-11-samitolvanen@google.com Signed-off-by: Palmer Dabbelt

riscv: Add types to indirectly called assembly functions

2023-08-23T21:16:37Z

With CONFIG_CFI_CLANG, assembly functions indirectly called from C code must be annotated with type identifiers to pass CFI checking. Use the SYM_TYPED_START macro to add types to the relevant functions. Reviewed-by: Kees Cook Tested-by: Nathan Chancellor Signed-off-by: Sami Tolvanen Link: https://lore.kernel.org/r/20230710183544.999540-10-samitolvanen@google.com Signed-off-by: Palmer Dabbelt

riscv: ftrace: Enable HAVE_FUNCTION_GRAPH_RETVAL

2023-06-22T14:39:56Z

The previous patch ("function_graph: Support recording and printing the return value of function") has laid the groundwork for the for the funcgraph-retval, and this modification makes it available on the RISC-V platform. We introduce a new structure called fgraph_ret_regs for the RISC-V platform to hold return registers and the frame pointer. We then fill its content in the return_to_handler and pass its address to the function ftrace_return_to_handler to record the return value. Link: https://lore.kernel.org/linux-trace-kernel/a8d71b12259f90e7e63d0ea654fcac95b0232bbc.1680954589.git.pengdonglin@sangfor.com.cn Signed-off-by: Donglin Peng Acked-by: Palmer Dabbelt Signed-off-by: Steven Rostedt (Google)

RISC-V: preserve a1 in mcount

2022-12-02T18:04:37Z

The RISC-V ELF psABI states that both a0 and a1 are used for return values so we should preserve them both in return_to_handler. This is especially important for RV32 for functions returning a 64-bit quantity otherwise the return value can be corrupted and undefined behaviour results. Reviewed-by: Andrew Jones Signed-off-by: Jamie Iles Link: https://lore.kernel.org/r/20221115200832.706370-4-jamie@jamieiles.com Signed-off-by: Palmer Dabbelt