bpf_encoder: Fix a verbose output issue
Yonghong Song
reviewerPATCH v2 1/1 bpf_encoder: Fix a verbose output issue
For the following test.c:
$ cat test.c
unsigned tar(int a);
__attribute__((noinline)) static int foo(int a, int b)
{
return tar(a) + tar(a + 1);
}
__attribute__((noinline)) int bar(int a)
{
foo(a, 1);
return 0;
}
The llvm compilation:
$ clang -O2 -g -c test.c
And then
$ pahole -JV test.o
btf_encoder__new: 'test.o' doesn't have '.data..percpu' sectio n
File test.o:
[1] INT unsigned int size=4 nr_bits=32 encoding=(none)
[2] INT int size=4 nr_bits=32 encoding=SIGNED
search cu 'test.c' for percpu global variables.
[3] FUNC_PROTO (anon) return=2 args=(2 a, [4] FUNC bar type_id=3
[5] FUNC_PROTO (anon) return=2 args=(2 a, 2 b, [6] FUNC foo type_id=5The above confused format is due to btf_encoder__add_func_proto_for_state(). The "is_last = param_idx == nr_params" is always false since param_idx starts from 0. The below change fixed the issue:
is_last = param_idx == (nr_params - 1)
With the fix, 'pahole -JV test.o' will produce the following:
...
[3] FUNC_PROTO (anon) return=2 args=(2 a)
[4] FUNC bar type_id=3
[5] FUNC_PROTO (anon) return=2 args=(2 a, 2 b)
[6] FUNC foo type_id=5
...
In addition, in btf_encoder__add_func_proto_for_ftype(), we have
++param_idx;
if (ftype->unspec_parms) { ... }
This is correct but it is misleading since '++param_idx' is only needed inside the above 'if' condition. So put '++param_idx' inside the 'if' condition to make code cleaner.
--- btf_encoder.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-)
btf_encoder.c +4 -3
@@ -865,10 +865,11 @@ static int32_t btf_encoder__add_func_proto_for_ftype(struct btf_encoder *encoderreturn -1 ; } - ++param_idx; -if (ftype->unspec_parms) +if (ftype->unspec_parms) { + ++param_idx;if (btf_encoder__add_func_param (encoder,NULL ,0 , param_idx == nr_params))return -1 ; + }return id; }
@@ -895,7 +896,7 @@ static int32_t btf_encoder__add_func_proto_for_state(struct btf_encoder *encoderfor (param_idx =0 ; param_idx < nr_params; param_idx++) { p = &state->parms[param_idx]; name =btf__name_by_offset (btf, p->name_off); - is_last = param_idx == nr_params; + is_last = param_idx == (nr_params -1 );/* adding BTF data may result in a move of the * name string memory, so make a temporary copy.
Signature · 1 line
2.47.3
Yonghong Song
reviewerQuoted 68 lines · Yonghong Song
For the following
test.c:$ cat test.c unsigned tar(int a); __attribute__((noinline)) static int foo(int a, int b) { return tar(a) + tar(a + 1); } __attribute__((noinline)) int bar(int a) { foo(a, 1); return 0; } The llvm compilation: $ clang -O2 -g -c test.c And then $ pahole -JV test.o btf_encoder__new: 'test.o' doesn't have '.data..percpu' sectio n File test.o: [1] INT unsigned int size=4 nr_bits=32 encoding=(none) [2] INT int size=4 nr_bits=32 encoding=SIGNED search cu 'test.c' for percpu global variables. [3] FUNC_PROTO (anon) return=2 args=(2 a, [4] FUNC bar type_id=3 [5] FUNC_PROTO (anon) return=2 args=(2 a, 2 b, [6] FUNC foo type_id=5The above confused format is due to
btf_encoder__add_func_proto_for_state(). The "is_last=param_idx==nr_params" is always false sinceparam_idxstarts from 0. The below change fixed the issue:is_last = param_idx == (nr_params - 1)With the fix, 'pahole -JV test.o' will produce the following:
... [3] FUNC_PROTO (anon) return=2 args=(2 a) [4] FUNC bar type_id=3 [5] FUNC_PROTO (anon) return=2 args=(2 a, 2 b) [6] FUNC foo type_id=5 ...In addition, in
btf_encoder__add_func_proto_for_ftype(), we have++param_idx; if (ftype->unspec_parms) { ... }This is correct but it is misleading since '++param_idx' is only needed inside the above 'if' condition. So put '++param_idx' inside the 'if' condition to make code cleaner.
Signed-off-by Yonghong Song <[email protected]>--- btf_encoder.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-)btf_encoder.c +3
-2@@ -865,10 +865,11 @@ static int32_t btf_encoder__add_func_proto_for_ftype(struct btf_encoder *encoderreturn -1 ; } - ++param_idx; -if (ftype->unspec_parms) +if (ftype->unspec_parms) { + ++param_idx;if (btf_encoder__add_func_param (encoder,NULL ,0 , param_idx == nr_params))return -1 ; + }return id; }
btf_encoder.c static int32_t btf_encoder__add_func_proto_for_state(struct btf_encoder *encoder · line 902
Quoted 8 lines
btf_encoder.c +1
-1@@ -895,7 +896,7 @@ static int32_t btf_encoder__add_func_proto_for_state(struct btf_encoder *encoderfor (param_idx =0 ; param_idx < nr_params; param_idx++) { p = &state->parms[param_idx]; name =btf__name_by_offset (btf, p->name_off); - is_last = param_idx == nr_params; + is_last = param_idx == (nr_params -1 );/* adding BTF data may result in a move of the * name string memory, so make a temporary copy.
Sorry. The tag should be '[PATCH dwarves v2]' instead of '[PATCH bpf-next v2]'.
Alan Maguire
AppliedQuoted 37 lines · Yonghong Song
For the following
test.c:$ cat test.c unsigned tar(int a); __attribute__((noinline)) static int foo(int a, int b) { return tar(a) + tar(a + 1); } __attribute__((noinline)) int bar(int a) { foo(a, 1); return 0; } The llvm compilation: $ clang -O2 -g -c test.c And then $ pahole -JV test.o btf_encoder__new: 'test.o' doesn't have '.data..percpu' sectio n File test.o: [1] INT unsigned int size=4 nr_bits=32 encoding=(none) [2] INT int size=4 nr_bits=32 encoding=SIGNED search cu 'test.c' for percpu global variables. [3] FUNC_PROTO (anon) return=2 args=(2 a, [4] FUNC bar type_id=3 [5] FUNC_PROTO (anon) return=2 args=(2 a, 2 b, [6] FUNC foo type_id=5The above confused format is due to
btf_encoder__add_func_proto_for_state(). The "is_last=param_idx==nr_params" is always false sinceparam_idxstarts from 0. The below change fixed the issue:
is_last=param_idx== (nr_params- 1)With the fix, 'pahole -JV test.o' will produce the following:
...
[3]FUNC_PROTO(anon) return=2 args=(2 a)
[4] FUNC bar type_id=3
[5]FUNC_PROTO(anon) return=2 args=(2 a, 2 b)
[6] FUNC foo type_id=5
...
Quoted 8 lines
In addition, in
btf_encoder__add_func_proto_for_ftype(), we have
++param_idx;
if (ftype->unspec_parms) { ... }
This is correct but it is misleading since '++param_idx' is only needed inside the above 'if' condition. So put '++param_idx' inside the 'if' condition to make code cleaner.Signed-off-by Yonghong Song <[email protected]>
applied, thanks!
Alan