[PATCH 4/8] arm64: module/ftrace: intialize PLT at load time

From: Mark Rutland
Date: Mon Oct 21 2019 - 12:35:14 EST


Currently we lazily-initialize a module's ftrace PLT at runtime when we
install the first ftrace call. To do so we have to apply a number of
sanity checks, transiently mark the module text as RW, and perform an
IPI as part of handling Neoverse-N1 erratum #1542419.

We only expect the ftrace trampoline to point at ftrace_caller() (AKA
FTRACE_ADDR), so let's simplify all of this by intializing the PLT at
module load time, before the module loader marks the module RO and
performs the intial I-cache maintenance for the module.

Thus we can rely on the module having been correctly intialized, and can
simplify the runtime work necessary to install an ftrace call in a
module. This will also allow for the removal of module_disable_ro().

Tested by forcing ftrace_make_call() to use the module PLT, and then
loading up a module after setting up ftrace with:

| echo ":mod:<module-name>" > set_ftrace_filter;
| echo function > current_tracer;
| modprobe <module-name>

Since FTRACE_ADDR is only defined when CONFIG_DYNAMIC_FTRACE is
selected, we wrap its use along with most of module_init_ftrace_plt()
with ifdeffery rather than using IS_ENABLED().

Signed-off-by: Mark Rutland <mark.rutland@xxxxxxx>
Cc: Ard Biesheuvel <ard.biesheuvel@xxxxxxxxxx>
Cc: Catalin Marinas <catalin.marinas@xxxxxxx>
Cc: James Morse <james.morse@xxxxxxx>
Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
Cc: Will Deacon <will@xxxxxxxxxx>
---
arch/arm64/kernel/ftrace.c | 55 ++++++++++++----------------------------------
arch/arm64/kernel/module.c | 32 +++++++++++++++++----------
2 files changed, 35 insertions(+), 52 deletions(-)

diff --git a/arch/arm64/kernel/ftrace.c b/arch/arm64/kernel/ftrace.c
index 06e56b470315..822718eafdb4 100644
--- a/arch/arm64/kernel/ftrace.c
+++ b/arch/arm64/kernel/ftrace.c
@@ -73,10 +73,22 @@ int ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr)

if (offset < -SZ_128M || offset >= SZ_128M) {
#ifdef CONFIG_ARM64_MODULE_PLTS
- struct plt_entry trampoline, *dst;
struct module *mod;

/*
+ * There is only one ftrace trampoline per module. For now,
+ * this is not a problem since on arm64, all dynamic ftrace
+ * invocations are routed via ftrace_caller(). This will need
+ * to be revisited if support for multiple ftrace entry points
+ * is added in the future, but for now, the pr_err() below
+ * deals with a theoretical issue only.
+ */
+ if (addr != FTRACE_ADDR) {
+ pr_err("ftrace: far branches to multiple entry points unsupported inside a single module\n");
+ return -EINVAL;
+ }
+
+ /*
* On kernels that support module PLTs, the offset between the
* branch instruction and its target may legally exceed the
* range of an ordinary relative 'bl' opcode. In this case, we
@@ -93,46 +105,7 @@ int ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr)
if (WARN_ON(!mod))
return -EINVAL;

- /*
- * There is only one ftrace trampoline per module. For now,
- * this is not a problem since on arm64, all dynamic ftrace
- * invocations are routed via ftrace_caller(). This will need
- * to be revisited if support for multiple ftrace entry points
- * is added in the future, but for now, the pr_err() below
- * deals with a theoretical issue only.
- *
- * Note that PLTs are place relative, and plt_entries_equal()
- * checks whether they point to the same target. Here, we need
- * to check if the actual opcodes are in fact identical,
- * regardless of the offset in memory so use memcmp() instead.
- */
- dst = mod->arch.ftrace_trampoline;
- trampoline = get_plt_entry(addr, dst);
- if (memcmp(dst, &trampoline, sizeof(trampoline))) {
- if (plt_entry_is_initialized(dst)) {
- pr_err("ftrace: far branches to multiple entry points unsupported inside a single module\n");
- return -EINVAL;
- }
-
- /* point the trampoline to our ftrace entry point */
- module_disable_ro(mod);
- *dst = trampoline;
- module_enable_ro(mod, true);
-
- /*
- * Ensure updated trampoline is visible to instruction
- * fetch before we patch in the branch. Although the
- * architecture doesn't require an IPI in this case,
- * Neoverse-N1 erratum #1542419 does require one
- * if the TLB maintenance in module_enable_ro() is
- * skipped due to rodata_enabled. It doesn't seem worth
- * it to make it conditional given that this is
- * certainly not a fast-path.
- */
- flush_icache_range((unsigned long)&dst[0],
- (unsigned long)&dst[1]);
- }
- addr = (unsigned long)dst;
+ addr = (unsigned long)mod->arch.ftrace_trampoline;
#else /* CONFIG_ARM64_MODULE_PLTS */
return -EINVAL;
#endif /* CONFIG_ARM64_MODULE_PLTS */
diff --git a/arch/arm64/kernel/module.c b/arch/arm64/kernel/module.c
index 763a86d52fef..5f5bc3b94da7 100644
--- a/arch/arm64/kernel/module.c
+++ b/arch/arm64/kernel/module.c
@@ -9,6 +9,7 @@

#include <linux/bitops.h>
#include <linux/elf.h>
+#include <linux/ftrace.h>
#include <linux/gfp.h>
#include <linux/kasan.h>
#include <linux/kernel.h>
@@ -485,24 +486,33 @@ static const Elf_Shdr *find_section(const Elf_Ehdr *hdr,
return NULL;
}

+int module_init_ftrace_plt(const Elf_Ehdr *hdr,
+ const Elf_Shdr *sechdrs,
+ struct module *mod)
+{
+#if defined(CONFIG_ARM64_MODULE_PLTS) && defined(CONFIG_DYNAMIC_FTRACE)
+ const Elf_Shdr *s;
+ struct plt_entry *plt;
+
+ s = find_section(hdr, sechdrs, ".text.ftrace_trampoline");
+ if (!s)
+ return -ENOEXEC;
+
+ plt = (void *)s->sh_addr;
+ *plt = get_plt_entry(FTRACE_ADDR, plt);
+ mod->arch.ftrace_trampoline = plt;
+#endif
+ return 0;
+}
+
int module_finalize(const Elf_Ehdr *hdr,
const Elf_Shdr *sechdrs,
struct module *me)
{
const Elf_Shdr *s;
-
s = find_section(hdr, sechdrs, ".altinstructions");
if (s)
apply_alternatives_module((void *)s->sh_addr, s->sh_size);

-#ifdef CONFIG_ARM64_MODULE_PLTS
- if (IS_ENABLED(CONFIG_DYNAMIC_FTRACE)) {
- s = find_section(hdr, sechdrs, ".text.ftrace_trampoline");
- if (!s)
- return -ENOEXEC;
- me->arch.ftrace_trampoline = (void *)s->sh_addr;
- }
-#endif
-
- return 0;
+ return module_init_ftrace_plt(hdr, sechdrs, me);
}
--
2.11.0