| Message ID | 20240730005433.3559731-4-song@kernel.org (mailing list archive) |
|---|---|
| State | Superseded |
| Headers | show |
| Series | Fix kallsyms with CONFIG_LTO_CLANG | expand |
On Mon, 29 Jul 2024 17:54:33 -0700 Song Liu <song@kernel.org> wrote: > Use the new kallsyms APIs that matches symbols name with .llvm.<hash> > suffix. This allows userspace tools to get kprobes on the expected > function name, while the actual symbol has a .llvm.<hash> suffix. > _kprobe_addr@kernel/kprobes.c may also fail with this change. Thanks, > This only effects kernel compared with CONFIG_LTO_CLANG. > > Signed-off-by: Song Liu <song@kernel.org> > --- > kernel/trace/trace_kprobe.c | 10 ++++++++-- > 1 file changed, 8 insertions(+), 2 deletions(-) > > diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c > index 61a6da808203..c319382c1a09 100644 > --- a/kernel/trace/trace_kprobe.c > +++ b/kernel/trace/trace_kprobe.c > @@ -202,7 +202,8 @@ unsigned long trace_kprobe_address(struct trace_kprobe *tk) > > if (tk->symbol) { > addr = (unsigned long) > - kallsyms_lookup_name(trace_kprobe_symbol(tk)); > + kallsyms_lookup_name_or_prefix(trace_kprobe_symbol(tk)); > + > if (addr) > addr += tk->rp.kp.offset; > } else { > @@ -766,8 +767,13 @@ static unsigned int number_of_same_symbols(const char *mod, const char *func_nam > { > struct sym_count_ctx ctx = { .count = 0, .name = func_name }; > > - if (!mod) > + if (!mod) { > kallsyms_on_each_match_symbol(count_symbols, func_name, &ctx.count); > + if (IS_ENABLED(CONFIG_LTO_CLANG) && !ctx.count) { > + kallsyms_on_each_match_symbol_or_prefix( > + count_symbols, func_name, &ctx.count); > + } > + } > > module_kallsyms_on_each_symbol(mod, count_mod_symbols, &ctx); > > -- > 2.43.0 >
Hi Masami, > On Jul 30, 2024, at 6:04 AM, Masami Hiramatsu <mhiramat@kernel.org> wrote: > > On Mon, 29 Jul 2024 17:54:33 -0700 > Song Liu <song@kernel.org> wrote: > >> Use the new kallsyms APIs that matches symbols name with .llvm.<hash> >> suffix. This allows userspace tools to get kprobes on the expected >> function name, while the actual symbol has a .llvm.<hash> suffix. >> > > _kprobe_addr@kernel/kprobes.c may also fail with this change. > Thanks for catching this! I will fix this in the next version. Song
diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index 61a6da808203..c319382c1a09 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -202,7 +202,8 @@ unsigned long trace_kprobe_address(struct trace_kprobe *tk) if (tk->symbol) { addr = (unsigned long) - kallsyms_lookup_name(trace_kprobe_symbol(tk)); + kallsyms_lookup_name_or_prefix(trace_kprobe_symbol(tk)); + if (addr) addr += tk->rp.kp.offset; } else { @@ -766,8 +767,13 @@ static unsigned int number_of_same_symbols(const char *mod, const char *func_nam { struct sym_count_ctx ctx = { .count = 0, .name = func_name }; - if (!mod) + if (!mod) { kallsyms_on_each_match_symbol(count_symbols, func_name, &ctx.count); + if (IS_ENABLED(CONFIG_LTO_CLANG) && !ctx.count) { + kallsyms_on_each_match_symbol_or_prefix( + count_symbols, func_name, &ctx.count); + } + } module_kallsyms_on_each_symbol(mod, count_mod_symbols, &ctx);
Use the new kallsyms APIs that matches symbols name with .llvm.<hash> suffix. This allows userspace tools to get kprobes on the expected function name, while the actual symbol has a .llvm.<hash> suffix. This only effects kernel compared with CONFIG_LTO_CLANG. Signed-off-by: Song Liu <song@kernel.org> --- kernel/trace/trace_kprobe.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-)