FTRACE's function tracer currently doesn't always work on ARM with
MODULE_PLT option enabled. If the module is loaded too far, FTRACE's
code modifier cannot cope with introduced veneers and turns the
function tracer off globally.
ARM64 already has a solution for the problem, refer to the following
patches:
arm64: ftrace: emit ftrace-mod.o contents through code
arm64: module-plts: factor out PLT generation code for ftrace
arm64: ftrace: fix !CONFIG_ARM64_MODULE_PLTS kernels
arm64: ftrace: fix building without CONFIG_MODULES
arm64: ftrace: add support for far branches to dynamic ftrace
arm64: ftrace: don't validate branch via PLT in ftrace_make_nop()
But the presented ARM variant has just a half of the footprint in terms of
the changed LoCs. It also retains the code validation-before-modification
instead of switching it off.
Changelog:
v5:
* BUILD_BUG_ON() ensures fixed_plts[] always fits one PLT block
* use "for" loop instead of "while"
* scripts/recordmcount is filtering reloc types
v4:
* Fixed build without CONFIG_FUNCTION_TRACER
* Reorganized pre-allocated PLTs handling in get_module_plt(),
now compiler eliminates the whole FTRACE-related handling code
if ARRAY_SIZE(fixed_plts) == 0
v3:
* Only extend struct dyn_arch_ftrace when ARM_MODULE_PLTS is enabled
v2:
* As suggested by Steven Rostedt, refrain from tree-wide API modification,
save module pointer in struct dyn_arch_ftrace instead (PowerPC way)
Alexander Sverdlin (3):
ARM: PLT: Move struct plt_entries definition to header
ARM: ftrace/recordmcount: filter relocation types
ARM: ftrace: Add MODULE_PLTS support
arch/arm/include/asm/ftrace.h | 3 +++
arch/arm/include/asm/module.h | 10 +++++++
arch/arm/kernel/ftrace.c | 62 ++++++++++++++++++++++++++++++++++++-------
arch/arm/kernel/module-plts.c | 53 ++++++++++++++++++++++++++++--------
scripts/recordmcount.c | 17 ++++++++++++
5 files changed, 125 insertions(+), 20 deletions(-)
--
2.4.6
If code in arch/arm/kernel/ftrace.c would operate on mcount() pointer the
following may be generated:
00000230 <prealloc_fixed_plts>:
230: b5f8 push {r3, r4, r5, r6, r7, lr}
232: b500 push {lr}
234: f7ff fffe bl 0 <__gnu_mcount_nc>
234: R_ARM_THM_CALL __gnu_mcount_nc
238: f240 0600 movw r6, #0
238: R_ARM_THM_MOVW_ABS_NC __gnu_mcount_nc
23c: f8d0 1180 ldr.w r1, [r0, #384] ; 0x180
FTRACE currently is not able to deal with it:
WARNING: CPU: 0 PID: 0 at .../kernel/trace/ftrace.c:1979 ftrace_bug+0x1ad/0x230()
...
CPU: 0 PID: 0 Comm: swapper/0 Not tainted 4.4.116-... #1
...
[<c0314e3d>] (unwind_backtrace) from [<c03115e9>] (show_stack+0x11/0x14)
[<c03115e9>] (show_stack) from [<c051a7f1>] (dump_stack+0x81/0xa8)
[<c051a7f1>] (dump_stack) from [<c0321c5d>] (warn_slowpath_common+0x69/0x90)
[<c0321c5d>] (warn_slowpath_common) from [<c0321cf3>] (warn_slowpath_null+0x17/0x1c)
[<c0321cf3>] (warn_slowpath_null) from [<c038ee9d>] (ftrace_bug+0x1ad/0x230)
[<c038ee9d>] (ftrace_bug) from [<c038f1f9>] (ftrace_process_locs+0x27d/0x444)
[<c038f1f9>] (ftrace_process_locs) from [<c08915bd>] (ftrace_init+0x91/0xe8)
[<c08915bd>] (ftrace_init) from [<c0885a67>] (start_kernel+0x34b/0x358)
[<c0885a67>] (start_kernel) from [<00308095>] (0x308095)
---[ end trace cb88537fdc8fa200 ]---
ftrace failed to modify [<c031266c>] prealloc_fixed_plts+0x8/0x60
actual: 44:f2:e1:36
ftrace record flags: 0
(0) expected tramp: c03143e9
Fix this by allowing only selected reloc types in __mcount_loc.
The list itself comes from the legacy recordmcount.pl script.
Signed-off-by: Alexander Sverdlin <[email protected]>
---
scripts/recordmcount.c | 17 +++++++++++++++++
1 file changed, 17 insertions(+)
diff --git a/scripts/recordmcount.c b/scripts/recordmcount.c
index 16e086d..a4888e9 100644
--- a/scripts/recordmcount.c
+++ b/scripts/recordmcount.c
@@ -53,6 +53,10 @@
#define R_AARCH64_ABS64 257
#endif
+#define R_ARM_PC24 1
+#define R_ARM_THM_CALL 10
+#define R_ARM_CALL 28
+
static int fd_map; /* File descriptor for file being modified. */
static int mmap_failed; /* Boolean flag. */
static char gpfx; /* prefix for global symbol name (sometimes '_') */
@@ -428,6 +432,18 @@ is_mcounted_section_name(char const *const txtname)
#define RECORD_MCOUNT_64
#include "recordmcount.h"
+static int arm_is_fake_mcount(Elf32_Rel const *rp)
+{
+ switch (ELF32_R_TYPE(w(rp->r_info))) {
+ case R_ARM_THM_CALL:
+ case R_ARM_CALL:
+ case R_ARM_PC24:
+ return 0;
+ }
+
+ return 1;
+}
+
/* 64-bit EM_MIPS has weird ELF64_Rela.r_info.
* http://techpubs.sgi.com/library/manuals/4000/007-4658-001/pdf/007-4658-001.pdf
* We interpret Table 29 Relocation Operation (Elf64_Rel, Elf64_Rela) [p.40]
@@ -529,6 +545,7 @@ do_file(char const *const fname)
altmcount = "__gnu_mcount_nc";
make_nop = make_nop_arm;
rel_type_nop = R_ARM_NONE;
+ is_fake_mcount32 = arm_is_fake_mcount;
break;
case EM_AARCH64:
reltype = R_AARCH64_ABS64;
--
2.4.6
Teach ftrace_make_call() and ftrace_make_nop() about PLTs.
Teach PLT code about FTRACE and all its callbacks.
Otherwise the following might happen:
------------[ cut here ]------------
WARNING: CPU: 14 PID: 2265 at .../arch/arm/kernel/insn.c:14 __arm_gen_branch+0x83/0x8c()
...
Hardware name: LSI Axxia AXM55XX
[<c0314a49>] (unwind_backtrace) from [<c03115e9>] (show_stack+0x11/0x14)
[<c03115e9>] (show_stack) from [<c0519f51>] (dump_stack+0x81/0xa8)
[<c0519f51>] (dump_stack) from [<c032185d>] (warn_slowpath_common+0x69/0x90)
[<c032185d>] (warn_slowpath_common) from [<c03218f3>] (warn_slowpath_null+0x17/0x1c)
[<c03218f3>] (warn_slowpath_null) from [<c03143cf>] (__arm_gen_branch+0x83/0x8c)
[<c03143cf>] (__arm_gen_branch) from [<c0314337>] (ftrace_make_nop+0xf/0x24)
[<c0314337>] (ftrace_make_nop) from [<c038ebcb>] (ftrace_process_locs+0x27b/0x3e8)
[<c038ebcb>] (ftrace_process_locs) from [<c0378d79>] (load_module+0x11e9/0x1a44)
[<c0378d79>] (load_module) from [<c037974d>] (SyS_finit_module+0x59/0x84)
[<c037974d>] (SyS_finit_module) from [<c030e981>] (ret_fast_syscall+0x1/0x18)
---[ end trace e1b64ced7a89adcc ]---
------------[ cut here ]------------
WARNING: CPU: 14 PID: 2265 at .../kernel/trace/ftrace.c:1979 ftrace_bug+0x1b1/0x234()
...
Hardware name: LSI Axxia AXM55XX
[<c0314a49>] (unwind_backtrace) from [<c03115e9>] (show_stack+0x11/0x14)
[<c03115e9>] (show_stack) from [<c0519f51>] (dump_stack+0x81/0xa8)
[<c0519f51>] (dump_stack) from [<c032185d>] (warn_slowpath_common+0x69/0x90)
[<c032185d>] (warn_slowpath_common) from [<c03218f3>] (warn_slowpath_null+0x17/0x1c)
[<c03218f3>] (warn_slowpath_null) from [<c038e87d>] (ftrace_bug+0x1b1/0x234)
[<c038e87d>] (ftrace_bug) from [<c038ebd5>] (ftrace_process_locs+0x285/0x3e8)
[<c038ebd5>] (ftrace_process_locs) from [<c0378d79>] (load_module+0x11e9/0x1a44)
[<c0378d79>] (load_module) from [<c037974d>] (SyS_finit_module+0x59/0x84)
[<c037974d>] (SyS_finit_module) from [<c030e981>] (ret_fast_syscall+0x1/0x18)
---[ end trace e1b64ced7a89adcd ]---
ftrace failed to modify [<e9ef7006>] 0xe9ef7006
actual: 02:f0:3b:fa
ftrace record flags: 0
(0) expected tramp: c0314265
Signed-off-by: Alexander Sverdlin <[email protected]>
---
arch/arm/include/asm/ftrace.h | 3 +++
arch/arm/include/asm/module.h | 1 +
arch/arm/kernel/ftrace.c | 62 ++++++++++++++++++++++++++++++++++++-------
arch/arm/kernel/module-plts.c | 48 ++++++++++++++++++++++++++++++---
4 files changed, 101 insertions(+), 13 deletions(-)
diff --git a/arch/arm/include/asm/ftrace.h b/arch/arm/include/asm/ftrace.h
index 9e842ff..faeb6b1 100644
--- a/arch/arm/include/asm/ftrace.h
+++ b/arch/arm/include/asm/ftrace.h
@@ -19,6 +19,9 @@ struct dyn_arch_ftrace {
#ifdef CONFIG_OLD_MCOUNT
bool old_mcount;
#endif
+#ifdef CONFIG_ARM_MODULE_PLTS
+ struct module *mod;
+#endif
};
static inline unsigned long ftrace_call_adjust(unsigned long addr)
diff --git a/arch/arm/include/asm/module.h b/arch/arm/include/asm/module.h
index 6996405..e3d7a51 100644
--- a/arch/arm/include/asm/module.h
+++ b/arch/arm/include/asm/module.h
@@ -30,6 +30,7 @@ struct plt_entries {
struct mod_plt_sec {
struct elf32_shdr *plt;
+ struct plt_entries *plt_ent;
int plt_count;
};
diff --git a/arch/arm/kernel/ftrace.c b/arch/arm/kernel/ftrace.c
index 5617932..b55355f 100644
--- a/arch/arm/kernel/ftrace.c
+++ b/arch/arm/kernel/ftrace.c
@@ -98,6 +98,19 @@ int ftrace_arch_code_modify_post_process(void)
static unsigned long ftrace_call_replace(unsigned long pc, unsigned long addr)
{
+ s32 offset = addr - pc;
+ s32 blim = 0xfe000008;
+ s32 flim = 0x02000004;
+
+ if (IS_ENABLED(CONFIG_THUMB2_KERNEL)) {
+ blim = 0xff000004;
+ flim = 0x01000002;
+ }
+
+ if (IS_ENABLED(CONFIG_ARM_MODULE_PLTS) &&
+ (offset < blim || offset > flim))
+ return 0;
+
return arm_gen_branch_link(pc, addr);
}
@@ -166,10 +179,22 @@ int ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr)
{
unsigned long new, old;
unsigned long ip = rec->ip;
+ unsigned long aaddr = adjust_address(rec, addr);
old = ftrace_nop_replace(rec);
- new = ftrace_call_replace(ip, adjust_address(rec, addr));
+ new = ftrace_call_replace(ip, aaddr);
+
+#ifdef CONFIG_ARM_MODULE_PLTS
+ if (!new) {
+ struct module *mod = rec->arch.mod;
+
+ if (mod) {
+ aaddr = get_module_plt(mod, ip, aaddr);
+ new = ftrace_call_replace(ip, aaddr);
+ }
+ }
+#endif
return ftrace_modify_code(rec->ip, old, new, true);
}
@@ -199,20 +224,39 @@ int ftrace_make_nop(struct module *mod,
unsigned long new;
int ret;
- old = ftrace_call_replace(ip, adjust_address(rec, addr));
- new = ftrace_nop_replace(rec);
- ret = ftrace_modify_code(ip, old, new, true);
+#ifdef CONFIG_ARM_MODULE_PLTS
+ /* mod is only supplied during module loading */
+ if (!mod)
+ mod = rec->arch.mod;
+ else
+ rec->arch.mod = mod;
+#endif
-#ifdef CONFIG_OLD_MCOUNT
- if (ret == -EINVAL && addr == MCOUNT_ADDR) {
- rec->arch.old_mcount = true;
+ for (;;) {
+ unsigned long aaddr = adjust_address(rec, addr);
+
+ old = ftrace_call_replace(ip, aaddr);
+
+#ifdef CONFIG_ARM_MODULE_PLTS
+ if (!old && mod) {
+ aaddr = get_module_plt(mod, ip, aaddr);
+ old = ftrace_call_replace(ip, aaddr);
+ }
+#endif
- old = ftrace_call_replace(ip, adjust_address(rec, addr));
new = ftrace_nop_replace(rec);
ret = ftrace_modify_code(ip, old, new, true);
- }
+
+#ifdef CONFIG_OLD_MCOUNT
+ if (ret == -EINVAL && !rec->arch.old_mcount) {
+ rec->arch.old_mcount = true;
+ continue;
+ }
#endif
+ break;
+ }
+
return ret;
}
diff --git a/arch/arm/kernel/module-plts.c b/arch/arm/kernel/module-plts.c
index f272711..da9a8a3 100644
--- a/arch/arm/kernel/module-plts.c
+++ b/arch/arm/kernel/module-plts.c
@@ -7,6 +7,7 @@
*/
#include <linux/elf.h>
+#include <linux/ftrace.h>
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/sort.h>
@@ -22,19 +23,56 @@
(PLT_ENT_STRIDE - 8))
#endif
+static const u32 fixed_plts[] = {
+#ifdef CONFIG_FUNCTION_TRACER
+ FTRACE_ADDR,
+ MCOUNT_ADDR,
+#endif
+#ifdef CONFIG_OLD_MCOUNT
+ (unsigned long)ftrace_caller_old,
+ (unsigned long)mcount,
+#endif
+};
+
static bool in_init(const struct module *mod, unsigned long loc)
{
return loc - (u32)mod->init_layout.base < mod->init_layout.size;
}
+static void prealloc_fixed(struct mod_plt_sec *pltsec, struct plt_entries *plt)
+{
+ int i;
+
+ if (!ARRAY_SIZE(fixed_plts) || pltsec->plt_count)
+ return;
+ pltsec->plt_count = ARRAY_SIZE(fixed_plts);
+
+ for (i = 0; i < ARRAY_SIZE(plt->ldr); ++i)
+ plt->ldr[i] = PLT_ENT_LDR;
+
+ BUILD_BUG_ON(sizeof(fixed_plts) > sizeof(plt->lit));
+ memcpy(plt->lit, fixed_plts, sizeof(fixed_plts));
+}
+
u32 get_module_plt(struct module *mod, unsigned long loc, Elf32_Addr val)
{
struct mod_plt_sec *pltsec = !in_init(mod, loc) ? &mod->arch.core :
&mod->arch.init;
+ struct plt_entries *plt;
+ int idx;
+
+ /* cache the address, ELF header is available only during module load */
+ if (!pltsec->plt_ent)
+ pltsec->plt_ent = (struct plt_entries *)pltsec->plt->sh_addr;
+ plt = pltsec->plt_ent;
- struct plt_entries *plt = (struct plt_entries *)pltsec->plt->sh_addr;
- int idx = 0;
+ prealloc_fixed(pltsec, plt);
+
+ for (idx = 0; idx < ARRAY_SIZE(fixed_plts); ++idx)
+ if (plt->lit[idx] == val)
+ return (u32)&plt->ldr[idx];
+ idx = 0;
/*
* Look for an existing entry pointing to 'val'. Given that the
* relocations are sorted, this will be the last entry we allocated.
@@ -182,8 +220,8 @@ static unsigned int count_plts(const Elf32_Sym *syms, Elf32_Addr base,
int module_frob_arch_sections(Elf_Ehdr *ehdr, Elf_Shdr *sechdrs,
char *secstrings, struct module *mod)
{
- unsigned long core_plts = 0;
- unsigned long init_plts = 0;
+ unsigned long core_plts = ARRAY_SIZE(fixed_plts);
+ unsigned long init_plts = ARRAY_SIZE(fixed_plts);
Elf32_Shdr *s, *sechdrs_end = sechdrs + ehdr->e_shnum;
Elf32_Sym *syms = NULL;
@@ -238,6 +276,7 @@ int module_frob_arch_sections(Elf_Ehdr *ehdr, Elf_Shdr *sechdrs,
mod->arch.core.plt->sh_size = round_up(core_plts * PLT_ENT_SIZE,
sizeof(struct plt_entries));
mod->arch.core.plt_count = 0;
+ mod->arch.core.plt_ent = NULL;
mod->arch.init.plt->sh_type = SHT_NOBITS;
mod->arch.init.plt->sh_flags = SHF_EXECINSTR | SHF_ALLOC;
@@ -245,6 +284,7 @@ int module_frob_arch_sections(Elf_Ehdr *ehdr, Elf_Shdr *sechdrs,
mod->arch.init.plt->sh_size = round_up(init_plts * PLT_ENT_SIZE,
sizeof(struct plt_entries));
mod->arch.init.plt_count = 0;
+ mod->arch.init.plt_ent = NULL;
pr_debug("%s: plt=%x, init.plt=%x\n", __func__,
mod->arch.core.plt->sh_size, mod->arch.init.plt->sh_size);
--
2.4.6
No functional change, later it will be re-used in several files.
Signed-off-by: Alexander Sverdlin <[email protected]>
---
arch/arm/include/asm/module.h | 9 +++++++++
arch/arm/kernel/module-plts.c | 9 ---------
2 files changed, 9 insertions(+), 9 deletions(-)
diff --git a/arch/arm/include/asm/module.h b/arch/arm/include/asm/module.h
index 89ad059..6996405 100644
--- a/arch/arm/include/asm/module.h
+++ b/arch/arm/include/asm/module.h
@@ -19,6 +19,15 @@ enum {
};
#endif
+#define PLT_ENT_STRIDE L1_CACHE_BYTES
+#define PLT_ENT_COUNT (PLT_ENT_STRIDE / sizeof(u32))
+#define PLT_ENT_SIZE (sizeof(struct plt_entries) / PLT_ENT_COUNT)
+
+struct plt_entries {
+ u32 ldr[PLT_ENT_COUNT];
+ u32 lit[PLT_ENT_COUNT];
+};
+
struct mod_plt_sec {
struct elf32_shdr *plt;
int plt_count;
diff --git a/arch/arm/kernel/module-plts.c b/arch/arm/kernel/module-plts.c
index 3d0c2e4..f272711 100644
--- a/arch/arm/kernel/module-plts.c
+++ b/arch/arm/kernel/module-plts.c
@@ -14,10 +14,6 @@
#include <asm/cache.h>
#include <asm/opcodes.h>
-#define PLT_ENT_STRIDE L1_CACHE_BYTES
-#define PLT_ENT_COUNT (PLT_ENT_STRIDE / sizeof(u32))
-#define PLT_ENT_SIZE (sizeof(struct plt_entries) / PLT_ENT_COUNT)
-
#ifdef CONFIG_THUMB2_KERNEL
#define PLT_ENT_LDR __opcode_to_mem_thumb32(0xf8dff000 | \
(PLT_ENT_STRIDE - 4))
@@ -26,11 +22,6 @@
(PLT_ENT_STRIDE - 8))
#endif
-struct plt_entries {
- u32 ldr[PLT_ENT_COUNT];
- u32 lit[PLT_ENT_COUNT];
-};
-
static bool in_init(const struct module *mod, unsigned long loc)
{
return loc - (u32)mod->init_layout.base < mod->init_layout.size;
--
2.4.6