This draft series adds PPC32 support to Sathvika's series.
Verified on pmac32 on QEMU.
It should in principle also work for PPC64 BE but for the time being
something goes wrong. In the beginning I had a segfaut hence the first
patch. But I still get no mcount section in the files.
Christophe Leroy (3):
objtool: Fix SEGFAULT
objtool: Use target file endianness instead of a compiled constant
objtool: Use target file class size instead of a compiled constant
Sathvika Vasireddy (4):
objtool: Add --mnop as an option to --mcount
objtool: Enable objtool to run only on files with ftrace enabled
objtool/powerpc: Enable objtool to be built on ppc
objtool/powerpc: Add --mcount specific implementation
Makefile | 4 +-
arch/powerpc/Kconfig | 2 +
arch/x86/Kconfig | 1 +
scripts/Makefile.build | 5 +-
tools/objtool/arch/powerpc/Build | 2 +
tools/objtool/arch/powerpc/decode.c | 88 +++++++++++++++++++
.../arch/powerpc/include/arch/cfi_regs.h | 11 +++
tools/objtool/arch/powerpc/include/arch/elf.h | 8 ++
.../arch/powerpc/include/arch/special.h | 21 +++++
tools/objtool/arch/powerpc/special.c | 19 ++++
.../arch/x86/include/arch/endianness.h | 9 --
tools/objtool/builtin-check.c | 14 +++
tools/objtool/check.c | 51 ++++++-----
tools/objtool/elf.c | 23 ++++-
tools/objtool/include/objtool/builtin.h | 1 +
tools/objtool/include/objtool/elf.h | 9 ++
tools/objtool/include/objtool/endianness.h | 29 +++---
tools/objtool/orc_dump.c | 11 ++-
tools/objtool/orc_gen.c | 4 +-
tools/objtool/special.c | 3 +-
20 files changed, 257 insertions(+), 58 deletions(-)
create mode 100644 tools/objtool/arch/powerpc/Build
create mode 100644 tools/objtool/arch/powerpc/decode.c
create mode 100644 tools/objtool/arch/powerpc/include/arch/cfi_regs.h
create mode 100644 tools/objtool/arch/powerpc/include/arch/elf.h
create mode 100644 tools/objtool/arch/powerpc/include/arch/special.h
create mode 100644 tools/objtool/arch/powerpc/special.c
delete mode 100644 tools/objtool/arch/x86/include/arch/endianness.h
--
2.35.3
Some architectures like powerpc support both endianness, it's
therefore not possible to fix the endianness via arch/endianness.h
because there is no easy way to get the target endianness at
build time.
Use the endianness recorded in the file objtool is working on.
Signed-off-by: Christophe Leroy <[email protected]>
---
.../arch/x86/include/arch/endianness.h | 9 ------
tools/objtool/check.c | 2 +-
tools/objtool/include/objtool/endianness.h | 29 +++++++++----------
tools/objtool/orc_dump.c | 11 +++++--
tools/objtool/orc_gen.c | 4 +--
tools/objtool/special.c | 3 +-
6 files changed, 27 insertions(+), 31 deletions(-)
delete mode 100644 tools/objtool/arch/x86/include/arch/endianness.h
diff --git a/tools/objtool/arch/x86/include/arch/endianness.h b/tools/objtool/arch/x86/include/arch/endianness.h
deleted file mode 100644
index 7c362527da20..000000000000
--- a/tools/objtool/arch/x86/include/arch/endianness.h
+++ /dev/null
@@ -1,9 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0-or-later */
-#ifndef _ARCH_ENDIANNESS_H
-#define _ARCH_ENDIANNESS_H
-
-#include <endian.h>
-
-#define __TARGET_BYTE_ORDER __LITTLE_ENDIAN
-
-#endif /* _ARCH_ENDIANNESS_H */
diff --git a/tools/objtool/check.c b/tools/objtool/check.c
index 6cb07e151588..cef1dd54d505 100644
--- a/tools/objtool/check.c
+++ b/tools/objtool/check.c
@@ -1971,7 +1971,7 @@ static int read_unwind_hints(struct objtool_file *file)
return -1;
}
- cfi.cfa.offset = bswap_if_needed(hint->sp_offset);
+ cfi.cfa.offset = bswap_if_needed(file->elf, hint->sp_offset);
cfi.type = hint->type;
cfi.end = hint->end;
diff --git a/tools/objtool/include/objtool/endianness.h b/tools/objtool/include/objtool/endianness.h
index 10241341eff3..ab0515ba0538 100644
--- a/tools/objtool/include/objtool/endianness.h
+++ b/tools/objtool/include/objtool/endianness.h
@@ -2,33 +2,30 @@
#ifndef _OBJTOOL_ENDIANNESS_H
#define _OBJTOOL_ENDIANNESS_H
-#include <arch/endianness.h>
#include <linux/kernel.h>
#include <endian.h>
-
-#ifndef __TARGET_BYTE_ORDER
-#error undefined arch __TARGET_BYTE_ORDER
-#endif
-
-#if __BYTE_ORDER != __TARGET_BYTE_ORDER
-#define __NEED_BSWAP 1
-#else
-#define __NEED_BSWAP 0
-#endif
+#include <objtool/elf.h>
/*
- * Does a byte swap if target endianness doesn't match the host, i.e. cross
+ * Does a byte swap if target file endianness doesn't match the host, i.e. cross
* compilation for little endian on big endian and vice versa.
* To be used for multi-byte values conversion, which are read from / about
* to be written to a target native endianness ELF file.
*/
-#define bswap_if_needed(val) \
+static inline bool need_bswap(struct elf *elf)
+{
+ return (__BYTE_ORDER == __LITTLE_ENDIAN) ^
+ (elf->ehdr.e_ident[EI_DATA] == ELFDATA2LSB);
+}
+
+#define bswap_if_needed(elf, val) \
({ \
__typeof__(val) __ret; \
+ bool __need_bswap = need_bswap(elf); \
switch (sizeof(val)) { \
- case 8: __ret = __NEED_BSWAP ? bswap_64(val) : (val); break; \
- case 4: __ret = __NEED_BSWAP ? bswap_32(val) : (val); break; \
- case 2: __ret = __NEED_BSWAP ? bswap_16(val) : (val); break; \
+ case 8: __ret = __need_bswap ? bswap_64(val) : (val); break; \
+ case 4: __ret = __need_bswap ? bswap_32(val) : (val); break; \
+ case 2: __ret = __need_bswap ? bswap_16(val) : (val); break; \
default: \
BUILD_BUG(); break; \
} \
diff --git a/tools/objtool/orc_dump.c b/tools/objtool/orc_dump.c
index f5a8508c42d6..4f1211fec82c 100644
--- a/tools/objtool/orc_dump.c
+++ b/tools/objtool/orc_dump.c
@@ -76,6 +76,7 @@ int orc_dump(const char *_objname)
GElf_Rela rela;
GElf_Sym sym;
Elf_Data *data, *symtab = NULL, *rela_orc_ip = NULL;
+ struct elf dummy_elf = {};
objname = _objname;
@@ -94,6 +95,12 @@ int orc_dump(const char *_objname)
return -1;
}
+ if (!elf64_getehdr(elf)) {
+ WARN_ELF("elf64_getehdr");
+ return -1;
+ }
+ memcpy(&dummy_elf.ehdr, elf64_getehdr(elf), sizeof(dummy_elf.ehdr));
+
if (elf_getshdrnum(elf, &nr_sections)) {
WARN_ELF("elf_getshdrnum");
return -1;
@@ -198,11 +205,11 @@ int orc_dump(const char *_objname)
printf(" sp:");
- print_reg(orc[i].sp_reg, bswap_if_needed(orc[i].sp_offset));
+ print_reg(orc[i].sp_reg, bswap_if_needed(&dummy_elf, orc[i].sp_offset));
printf(" bp:");
- print_reg(orc[i].bp_reg, bswap_if_needed(orc[i].bp_offset));
+ print_reg(orc[i].bp_reg, bswap_if_needed(&dummy_elf, orc[i].bp_offset));
printf(" type:%s end:%d\n",
orc_type_name(orc[i].type), orc[i].end);
diff --git a/tools/objtool/orc_gen.c b/tools/objtool/orc_gen.c
index dd3c64af9db2..1f22b7ebae58 100644
--- a/tools/objtool/orc_gen.c
+++ b/tools/objtool/orc_gen.c
@@ -97,8 +97,8 @@ static int write_orc_entry(struct elf *elf, struct section *orc_sec,
/* populate ORC data */
orc = (struct orc_entry *)orc_sec->data->d_buf + idx;
memcpy(orc, o, sizeof(*orc));
- orc->sp_offset = bswap_if_needed(orc->sp_offset);
- orc->bp_offset = bswap_if_needed(orc->bp_offset);
+ orc->sp_offset = bswap_if_needed(elf, orc->sp_offset);
+ orc->bp_offset = bswap_if_needed(elf, orc->bp_offset);
/* populate reloc for ip */
if (elf_add_reloc_to_insn(elf, ip_sec, idx * sizeof(int), R_X86_64_PC32,
diff --git a/tools/objtool/special.c b/tools/objtool/special.c
index e2223dd91c37..9c8d827f69af 100644
--- a/tools/objtool/special.c
+++ b/tools/objtool/special.c
@@ -87,7 +87,8 @@ static int get_alt_entry(struct elf *elf, struct special_entry *entry,
if (entry->feature) {
unsigned short feature;
- feature = bswap_if_needed(*(unsigned short *)(sec->data->d_buf +
+ feature = bswap_if_needed(elf,
+ *(unsigned short *)(sec->data->d_buf +
offset +
entry->feature));
arch_handle_alternative(feature, alt);
--
2.35.3
From: Sathvika Vasireddy <[email protected]>
This patch enables objtool --mcount on powerpc, and
adds implementation specific to powerpc.
Signed-off-by: Sathvika Vasireddy <[email protected]>
Signed-off-by: Christophe Leroy <[email protected]>
---
arch/powerpc/Kconfig | 1 +
tools/objtool/arch/powerpc/decode.c | 14 ++++++++++++++
tools/objtool/check.c | 12 +++++++-----
tools/objtool/elf.c | 15 +++++++++++++++
tools/objtool/include/objtool/elf.h | 1 +
5 files changed, 38 insertions(+), 5 deletions(-)
diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig
index 7c01229dd2e3..5ef8bf8eb202 100644
--- a/arch/powerpc/Kconfig
+++ b/arch/powerpc/Kconfig
@@ -233,6 +233,7 @@ config PPC
select HAVE_NMI if PERF_EVENTS || (PPC64 && PPC_BOOK3S)
select HAVE_OPTPROBES
select HAVE_OBJTOOL
+ select HAVE_OBJTOOL_MCOUNT if HAVE_OBJTOOL
select HAVE_PERF_EVENTS
select HAVE_PERF_EVENTS_NMI if PPC64
select HAVE_PERF_REGS
diff --git a/tools/objtool/arch/powerpc/decode.c b/tools/objtool/arch/powerpc/decode.c
index eb1542d155fe..048bb4cd2838 100644
--- a/tools/objtool/arch/powerpc/decode.c
+++ b/tools/objtool/arch/powerpc/decode.c
@@ -41,12 +41,26 @@ int arch_decode_instruction(struct objtool_file *file, const struct section *sec
struct list_head *ops_list)
{
u32 insn;
+ unsigned int opcode;
*immediate = 0;
insn = bswap_if_needed(file->elf, *(u32 *)(sec->data->d_buf + offset));
*len = 4;
*type = INSN_OTHER;
+ opcode = (insn >> 26);
+
+ switch (opcode) {
+ case 18: /* bl */
+ if ((insn & 3) == 1) {
+ *type = INSN_CALL;
+ *immediate = insn & 0x3fffffc;
+ if (*immediate & 0x2000000)
+ *immediate -= 0x4000000;
+ }
+ break;
+ }
+
return 0;
}
diff --git a/tools/objtool/check.c b/tools/objtool/check.c
index 7f0dc504dd92..70be5a72e838 100644
--- a/tools/objtool/check.c
+++ b/tools/objtool/check.c
@@ -834,7 +834,7 @@ static int create_mcount_loc_sections(struct objtool_file *file)
memset(loc, 0, size);
if (elf_add_reloc_to_insn(file->elf, sec, idx,
- R_X86_64_64,
+ elf_reloc_type_long(file->elf),
insn->sec, insn->offset))
return -1;
@@ -2185,7 +2185,7 @@ static int classify_symbols(struct objtool_file *file)
if (arch_is_retpoline(func))
func->retpoline_thunk = true;
- if (!strcmp(func->name, "__fentry__"))
+ if ((!strcmp(func->name, "__fentry__")) || (!strcmp(func->name, "_mcount")))
func->fentry = true;
if (is_profiling_func(func->name))
@@ -2261,9 +2261,11 @@ static int decode_sections(struct objtool_file *file)
* Must be before add_jump_destinations(), which depends on 'func'
* being set for alternatives, to enable proper sibling call detection.
*/
- ret = add_special_section_alts(file);
- if (ret)
- return ret;
+ if (opts.stackval || opts.orc || opts.uaccess || opts.noinstr) {
+ ret = add_special_section_alts(file);
+ if (ret)
+ return ret;
+ }
ret = add_jump_destinations(file);
if (ret)
diff --git a/tools/objtool/elf.c b/tools/objtool/elf.c
index 63218f5799c2..34b1c6817a5e 100644
--- a/tools/objtool/elf.c
+++ b/tools/objtool/elf.c
@@ -793,6 +793,21 @@ elf_create_section_symbol(struct elf *elf, struct section *sec)
return sym;
}
+int elf_reloc_type_long(struct elf *elf)
+{
+ switch (elf->ehdr.e_machine) {
+ case EM_X86_64:
+ return R_X86_64_64;
+ case EM_PPC64:
+ return R_PPC64_ADDR64;
+ case EM_PPC:
+ return R_PPC_ADDR32;
+ default:
+ WARN("unknown machine...");
+ exit(-1);
+ }
+}
+
int elf_add_reloc_to_insn(struct elf *elf, struct section *sec,
unsigned long offset, unsigned int type,
struct section *insn_sec, unsigned long insn_off)
diff --git a/tools/objtool/include/objtool/elf.h b/tools/objtool/include/objtool/elf.h
index c720c4476828..d10f4701715b 100644
--- a/tools/objtool/include/objtool/elf.h
+++ b/tools/objtool/include/objtool/elf.h
@@ -152,6 +152,7 @@ static inline int elf_class_size(struct elf *elf)
struct elf *elf_open_read(const char *name, int flags);
struct section *elf_create_section(struct elf *elf, const char *name, unsigned int sh_flags, size_t entsize, int nr);
+int elf_reloc_type_long(struct elf *elf);
int elf_add_reloc(struct elf *elf, struct section *sec, unsigned long offset,
unsigned int type, struct symbol *sym, s64 addend);
int elf_add_reloc_to_insn(struct elf *elf, struct section *sec,
--
2.35.3
On Wed, May 25, 2022 at 03:44:04PM +0530, Sathvika Vasireddy wrote:
> On 24/05/22 18:47, Christophe Leroy wrote:
> >This draft series adds PPC32 support to Sathvika's series.
> >Verified on pmac32 on QEMU.
> >
> >It should in principle also work for PPC64 BE but for the time being
> >something goes wrong. In the beginning I had a segfaut hence the first
> >patch. But I still get no mcount section in the files.
> Since PPC64 BE uses older elfv1 ABI, it prepends a dot to symbols.
> And so, the relocation records in case of PPC64BE point to "._mcount",
> rather than just "_mcount". We should be looking for "._mcount" to be
> able to generate mcount_loc section in the files.
The dotted symbol is on the actual function. The "normal" symbol is on
the "official procedure descriptor" (opd), which is what you get if you
(in C) take the address of a function. A procedure descriptor holds one
or two more pointers, the GOT and environment pointers. We don't use
the environment one, but the GOT pointer is necessary everywhere :-)
Segher
Hi Sathvika,
Le 25/05/2022 à 12:14, Sathvika Vasireddy a écrit :
> Hi Christophe,
>
> On 24/05/22 18:47, Christophe Leroy wrote:
>> This draft series adds PPC32 support to Sathvika's series.
>> Verified on pmac32 on QEMU.
>>
>> It should in principle also work for PPC64 BE but for the time being
>> something goes wrong. In the beginning I had a segfaut hence the first
>> patch. But I still get no mcount section in the files.
> Since PPC64 BE uses older elfv1 ABI, it prepends a dot to symbols.
> And so, the relocation records in case of PPC64BE point to "._mcount",
> rather than just "_mcount". We should be looking for "._mcount" to be
> able to generate mcount_loc section in the files.
>
> Like:
>
> diff --git a/tools/objtool/check.c b/tools/objtool/check.c
> index 70be5a72e838..7da5bf8c7236 100644
> --- a/tools/objtool/check.c
> +++ b/tools/objtool/check.c
> @@ -2185,7 +2185,7 @@ static int classify_symbols(struct objtool_file
> *file)
> if (arch_is_retpoline(func))
> func->retpoline_thunk = true;
>
> - if ((!strcmp(func->name, "__fentry__")) ||
> (!strcmp(func->name, "_mcount")))
> + if ((!strcmp(func->name, "__fentry__")) ||
> (!strcmp(func->name, "_mcount")) || (!strcmp(func->name, "._mcount")))
> func->fentry = true;
>
> if (is_profiling_func(func->name))
>
>
> With this change, I could see __mcount_loc section being
> generated in individual ppc64be object files.
>
Or should we implement an equivalent of arch_ftrace_match_adjust() in
objtool ?
Christophe
On 25/05/22 23:09, Christophe Leroy wrote:
> Hi Sathvika,
>
> Le 25/05/2022 à 12:14, Sathvika Vasireddy a écrit :
>> Hi Christophe,
>>
>> On 24/05/22 18:47, Christophe Leroy wrote:
>>> This draft series adds PPC32 support to Sathvika's series.
>>> Verified on pmac32 on QEMU.
>>>
>>> It should in principle also work for PPC64 BE but for the time being
>>> something goes wrong. In the beginning I had a segfaut hence the first
>>> patch. But I still get no mcount section in the files.
>> Since PPC64 BE uses older elfv1 ABI, it prepends a dot to symbols.
>> And so, the relocation records in case of PPC64BE point to "._mcount",
>> rather than just "_mcount". We should be looking for "._mcount" to be
>> able to generate mcount_loc section in the files.
>>
>> Like:
>>
>> diff --git a/tools/objtool/check.c b/tools/objtool/check.c
>> index 70be5a72e838..7da5bf8c7236 100644
>> --- a/tools/objtool/check.c
>> +++ b/tools/objtool/check.c
>> @@ -2185,7 +2185,7 @@ static int classify_symbols(struct objtool_file
>> *file)
>> if (arch_is_retpoline(func))
>> func->retpoline_thunk = true;
>>
>> - if ((!strcmp(func->name, "__fentry__")) ||
>> (!strcmp(func->name, "_mcount")))
>> + if ((!strcmp(func->name, "__fentry__")) ||
>> (!strcmp(func->name, "_mcount")) || (!strcmp(func->name, "._mcount")))
>> func->fentry = true;
>>
>> if (is_profiling_func(func->name))
>>
>>
>> With this change, I could see __mcount_loc section being
>> generated in individual ppc64be object files.
>>
> Or should we implement an equivalent of arch_ftrace_match_adjust() in
> objtool ?
Yeah, I think it makes more sense if we make it arch specific.
Thanks for the suggestion. I'll make this change in next revision :-)
- Sathvika
Hi Christophe,
On 24/05/22 18:47, Christophe Leroy wrote:
> This draft series adds PPC32 support to Sathvika's series.
> Verified on pmac32 on QEMU.
>
> It should in principle also work for PPC64 BE but for the time being
> something goes wrong. In the beginning I had a segfaut hence the first
> patch. But I still get no mcount section in the files.
Since PPC64 BE uses older elfv1 ABI, it prepends a dot to symbols.
And so, the relocation records in case of PPC64BE point to "._mcount",
rather than just "_mcount". We should be looking for "._mcount" to be
able to generate mcount_loc section in the files.
Like:
diff --git a/tools/objtool/check.c b/tools/objtool/check.c
index 70be5a72e838..7da5bf8c7236 100644
--- a/tools/objtool/check.c
+++ b/tools/objtool/check.c
@@ -2185,7 +2185,7 @@ static int classify_symbols(struct objtool_file *file)
if (arch_is_retpoline(func))
func->retpoline_thunk = true;
- if ((!strcmp(func->name, "__fentry__")) ||
(!strcmp(func->name, "_mcount")))
+ if ((!strcmp(func->name, "__fentry__")) ||
(!strcmp(func->name, "_mcount")) || (!strcmp(func->name, "._mcount")))
func->fentry = true;
if (is_profiling_func(func->name))
With this change, I could see __mcount_loc section being
generated in individual ppc64be object files.
- Sathvika
Le 25/05/2022 à 20:12, Sathvika Vasireddy a écrit :
>
> On 25/05/22 23:09, Christophe Leroy wrote:
>> Hi Sathvika,
>>
>> Le 25/05/2022 à 12:14, Sathvika Vasireddy a écrit :
>>> Hi Christophe,
>>>
>>> On 24/05/22 18:47, Christophe Leroy wrote:
>>>> This draft series adds PPC32 support to Sathvika's series.
>>>> Verified on pmac32 on QEMU.
>>>>
>>>> It should in principle also work for PPC64 BE but for the time being
>>>> something goes wrong. In the beginning I had a segfaut hence the first
>>>> patch. But I still get no mcount section in the files.
>>> Since PPC64 BE uses older elfv1 ABI, it prepends a dot to symbols.
>>> And so, the relocation records in case of PPC64BE point to "._mcount",
>>> rather than just "_mcount". We should be looking for "._mcount" to be
>>> able to generate mcount_loc section in the files.
>>>
>>> Like:
>>>
>>> diff --git a/tools/objtool/check.c b/tools/objtool/check.c
>>> index 70be5a72e838..7da5bf8c7236 100644
>>> --- a/tools/objtool/check.c
>>> +++ b/tools/objtool/check.c
>>> @@ -2185,7 +2185,7 @@ static int classify_symbols(struct objtool_file
>>> *file)
>>> if (arch_is_retpoline(func))
>>> func->retpoline_thunk = true;
>>>
>>> - if ((!strcmp(func->name, "__fentry__")) ||
>>> (!strcmp(func->name, "_mcount")))
>>> + if ((!strcmp(func->name, "__fentry__")) ||
>>> (!strcmp(func->name, "_mcount")) || (!strcmp(func->name, "._mcount")))
>>> func->fentry = true;
>>>
>>> if (is_profiling_func(func->name))
>>>
>>>
>>> With this change, I could see __mcount_loc section being
>>> generated in individual ppc64be object files.
>>>
>> Or should we implement an equivalent of arch_ftrace_match_adjust() in
>> objtool ?
>
> Yeah, I think it makes more sense if we make it arch specific.
> Thanks for the suggestion. I'll make this change in next revision :-)
>
Do you have any idea when you plan to send next revision ?
I'm really looking forward to submitting the inline static calls on top
of your series.
Thanks
Christophe
Hi Christophe,
On 15/06/22 21:33, Christophe Leroy wrote:
> Do you have any idea when you plan to send next revision ?
>
> I'm really looking forward to submitting the inline static calls on top
> of your series.
I'm planning to send RFC v3 next week.
- Sathvika
Hi Sathvika
Le 18/06/2022 à 06:56, Sathvika Vasireddy a écrit :
> Hi Christophe,
>
> On 15/06/22 21:33, Christophe Leroy wrote:
>> Do you have any idea when you plan to send next revision ?
>>
>> I'm really looking forward to submitting the inline static calls on top
>> of your series.
>
> I'm planning to send RFC v3 next week.
>
Is everything going ok ? Don't hesitate if you need help.
Christophe
Hi Christophe,
On 24/06/22 12:38, Christophe Leroy wrote:
> Is everything going ok ? Don't hesitate if you need help.
Yeah, sure. Thanks!
I just posted RFC v3 here:
https://patchwork.ozlabs.org/project/linuxppc-dev/cover/[email protected]/