Skip to contentSkip to conversation

perf symbol: Fix symbol loading from libbfd

PATCH Superseded Namhyung Kim 3 participants 16 messages Oct 12 perf events

v2 of 3 Superseded Latest is v3 →

lore
Share & tools
lore

Namhyung Kim

author maintainer
#0 lore raw

PATCH v2 1/4 perf symbol: Fix symbol loading from libbfd

The dso__load() should return a positive number (number of symbols loaded) when it succeded to load symbols. But dso__load_bfd_symbols() only returns 0 and it's hard to check if it actually loaded symbols or not. Let's return the value of dso__has_symbol() and check it in dso__load() so that it can return 1 for success.

Fixes eac9a4342e5447cad ("perf symbols: Try reading the symbol table with libbfd")
---
 tools/perf/tests/pe-file-parsing.c | 2 +-
 tools/perf/util/libbfd.c           | 2 +-
 tools/perf/util/symbol.c           | 4 ++--
 3 files changed, 4 insertions(+), 4 deletions(-)
tools/perf/tests/pe-file-parsing.c +1 -1
index 30c7da79e109b480..3979f38bf127c218 100644
@@ -59,7 +59,7 @@ static int run_dir(const char *d)
 	TEST_ASSERT_VAL("Failed to get dso", dso);
 
 	ret = dso__load_bfd_symbols(dso, debugfile);
-	TEST_ASSERT_VAL("Failed to load symbols", ret == 0);
+	TEST_ASSERT_VAL("Failed to load symbols", ret > 0);
 
 	dso__sort_by_name(dso);
 	sym = dso__find_symbol_by_name(dso, "main", &idx);
tools/perf/util/libbfd.c +1 -1
index a907d7c1ac948808..a4dcb406c79e9cee 100644
@@ -463,7 +463,7 @@ int dso__load_bfd_symbols(struct dso *dso, const char *debugfile)
 	symbols__fixup_duplicate(dso__symbols(dso));
 	dso__set_adjust_symbols(dso, true);
 
-	err = 0;
+	err = dso__has_symbols(dso);
 out_free:
 	free(symbols);
 out_close:
tools/perf/util/symbol.c +2 -2
index 5d98888d068c741a..a394e742f9b18812 100644
@@ -1930,8 +1930,8 @@ int dso__load(struct dso *dso, struct map *map)
 		if (nsexit)
 			nsinfo__mountns_enter(dso__nsinfo(dso), &nsc);
 
-		if (bfdrc == 0) {
-			ret = 0;
+		if (bfdrc > 0) {
+			ret = 1;
 			break;
 		}
 
Signature · 1 line
2.55.0
author maintainer
#1 lore raw

PATCH v2 2/4 perf symbol: Add missing symsrc__destroy() for libbfd

When libbfd loaded symbols and there's a pending syms_ss, it'd be lost or overwrite existing symbols. Let's destroy it before breaking the loop so that it can safely return from the function.

Fixes eac9a4342e5447cad ("perf symbols: Try reading the symbol table with libbfd")
---
 tools/perf/util/symbol.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
tools/perf/util/symbol.c +2 -1
index a394e742f9b18812..55a4b59195a08be5 100644
@@ -1932,7 +1932,7 @@ int dso__load(struct dso *dso, struct map *map)
 
 		if (bfdrc > 0) {
 			ret = 1;
-			break;
+			goto out_destroy;
 		}
 
 		if (!is_reg || sirc < 0)
@@ -1997,6 +1997,7 @@ int dso__load(struct dso *dso, struct map *map)
 			ret += nr_plt;
 	}
 
+out_destroy:
 	for (; ss_pos > 0; ss_pos--)
 		symsrc__destroy(&ss_[ss_pos - 1]);
 out_free:
Signature · 1 line
2.55.0
#2 lore raw

Re: [PATCH v2 2/4] perf symbol: Add missing symsrc__destroy() for libbfd

Sashiko has reviewed this patch and found no issues. It looks great!
Quoted 8 lines

When libbfd loaded symbols and there's a pending syms_ss, it'd be lost or overwrite existing symbols. Let's destroy it before breaking the loop so that it can safely return from the function.

Fixes eac9a4342e5447cad ("perf symbols: Try reading the symbol table with libbfd")

Sashiko has reviewed this patch and found no issues. It looks great!

Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=2
reviewer Reviewed-by on v2 2/4
#3 lore raw

Re: [PATCH v2 2/4] perf symbol: Add missing symsrc__destroy() for libbfd

In reply to Namhyung Kim
Quoted 9 lines · Namhyung Kim

When libbfd loaded symbols and there's a pending syms_ss, it'd be lost or overwrite existing symbols. Let's destroy it before breaking the loop so that it can safely return from the function.

Fixes eac9a4342e5447cad ("perf symbols: Try reading the symbol table with libbfd")

Thanks!
Ian

Quoted 28 lines
---
 tools/perf/util/symbol.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
tools/perf/util/symbol.c +2 -2
index a394e742f9b18812..55a4b59195a08be5 100644
@@ -1932,7 +1932,7 @@ int dso__load(struct dso *dso, struct map *map)
 
                if (bfdrc > 0) {
                        ret = 1;
-                       break;
+                       goto out_destroy;
                }
 
                if (!is_reg || sirc < 0)
@@ -1997,6 +1997,7 @@ int dso__load(struct dso *dso, struct map *map)
                        ret += nr_plt;
        }
 
+out_destroy:
        for (; ss_pos > 0; ss_pos--)
                symsrc__destroy(&ss_[ss_pos - 1]);
 out_free:
--

2.55.0

author maintainer
#4 lore raw

PATCH v2 3/4 perf symbol: Set dso symtab type for libbfd

In reply to Namhyung Kim

It missed to set the dso symtab type when the symbols are loaded by libbfd. When it happens both syms_ss and runtime_ss may not be set and won't call dso__load_sym().

---
 tools/perf/util/symbol.c | 1 +
 1 file changed, 1 insertion(+)
tools/perf/util/symbol.c +1 -0
index 55a4b59195a08be5..b63fa3c03f0e732c 100644
@@ -1931,6 +1931,7 @@ int dso__load(struct dso *dso, struct map *map)
 			nsinfo__mountns_enter(dso__nsinfo(dso), &nsc);
 
 		if (bfdrc > 0) {
+			dso__set_symtab_type(dso, symtab_type);
 			ret = 1;
 			goto out_destroy;
 		}
Signature · 1 line
2.55.0
#5 lore raw

Re: [PATCH v2 3/4] perf symbol: Set dso symtab type for libbfd

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider: - [Medium] perf symbol: incomplete dso state initialization for libbfd causing annotate/unwind failure --

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] perf symbol: incomplete dso state initialization for libbfd causing annotate/unwind failure --

commit 5475c2801d673496a1fee31f44ca921eca39f20c
Author: Namhyung Kim <[email protected]>

perf symbol: Set dso symtab type for libbfd

It missed to set the dso symtab type when the symbols are loaded by libbfd. When it happens both syms_ss and runtime_ss may not be set and won't call dso__load_sym().

tools/perf/util/symbol.c int dso__load(struct dso *dso, struct map *map) · line 1928

Quoted 9 lines
tools/perf/util/symbol.c +1 -0
index 4878d4292eb85..a90ce714b174b 100644
@@ -1925,6 +1925,7 @@ int dso__load(struct dso *dso, struct map *map)
 			nsinfo__mountns_enter(dso__nsinfo(dso), &nsc);
 
 		if (bfdrc > 0) {
+			dso__set_symtab_type(dso, symtab_type);

[Severity: Medium]
When dso__load() takes this early exit to out_destroy, it bypasses the rest of the loop which normally initializes symsrc_filename and binary_type:

		if (!syms_ss && symsrc__has_symtab(ss)) {
			syms_ss = ss;
			next_slot = true;
			if (!dso__symsrc_filename(dso))
				dso__set_symsrc_filename(dso, strdup(name));
		}
...
		if (next_slot) {
...
			if (dso__binary_type(dso) == DSO_BINARY_TYPE__NOT_FOUND || ... )
				dso__set_binary_type(dso, symtab_type);

Does skipping these initializations leave the DSO's binary_type as DSO_BINARY_TYPE__NOT_FOUND?

If so, could this cause perf annotate to fail later on, since symbol__disassemble() checks for this exact condition and returns an error?

tools/perf/util/disasm.c:symbol__disassemble()
...

	if (dso__binary_type(dso) == DSO_BINARY_TYPE__NOT_FOUND) {
		return SYMBOL_ANNOTATE_ERRNO__COULDNT_DETERMINE_FILE_TYPE;
	}

...

Quoted 3 lines
tools/perf/util/symbol.c +0 -0
 			ret = 1;
 			goto out_destroy;
 		}
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=3
author maintainer
#6 lore raw

Re: [PATCH v2 3/4] perf symbol: Set dso symtab type for libbfd

Quoted 39 lines · On Mon, Oct 12, 2026 at 05:40:08AM +0000, [email protected] wrote:

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] perf symbol: incomplete dso state initialization for libbfd causing annotate/unwind failure --

commit 5475c2801d673496a1fee31f44ca921eca39f20c
Author: Namhyung Kim <[email protected]>

perf symbol: Set dso symtab type for libbfd

It missed to set the dso symtab type when the symbols are loaded by libbfd. When it happens both syms_ss and runtime_ss may not be set and won't call dso__load_sym().

tools/perf/util/symbol.c int dso__load(struct dso *dso, struct map *map) · line 1928

Quoted 9 lines
tools/perf/util/symbol.c +1 -0
index 4878d4292eb85..a90ce714b174b 100644
@@ -1925,6 +1925,7 @@ int dso__load(struct dso *dso, struct map *map)
 			nsinfo__mountns_enter(dso__nsinfo(dso), &nsc);
 
 		if (bfdrc > 0) {
+			dso__set_symtab_type(dso, symtab_type);

[Severity: Medium]
When dso__load() takes this early exit to out_destroy, it bypasses the rest of the loop which normally initializes symsrc_filename and binary_type:

		if (!syms_ss && symsrc__has_symtab(ss)) {
			syms_ss = ss;
			next_slot = true;
			if (!dso__symsrc_filename(dso))
				dso__set_symsrc_filename(dso, strdup(name));
		}
...
		if (next_slot) {
...
			if (dso__binary_type(dso) == DSO_BINARY_TYPE__NOT_FOUND || ... )
				dso__set_binary_type(dso, symtab_type);
Quoted 5 lines

Does skipping these initializations leave the DSO's binary_type as DSO_BINARY_TYPE__NOT_FOUND?

If so, could this cause perf annotate to fail later on, since symbol__disassemble() checks for this exact condition and returns an error?

This is what I'm working on in another patchset. It'd be totally fine not to set symsrc_filename and binary_type with that change.

Thanks,
Namhyung

Quoted 14 lines

tools/perf/util/disasm.c:symbol__disassemble()
...

	if (dso__binary_type(dso) == DSO_BINARY_TYPE__NOT_FOUND) {
		return SYMBOL_ANNOTATE_ERRNO__COULDNT_DETERMINE_FILE_TYPE;
	}

...

Quoted 3 lines
tools/perf/util/symbol.c +0 -0
 			ret = 1;
 			goto out_destroy;
 		}
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=3
author maintainer
#7 lore raw

PATCH v2 4/4 perf test: Update PE file test for other locations

In reply to Namhyung Kim

The test__pe_file_parsing() expects the test is in the tools/perf directory, but it could be root or a separate build directory. Let's update the logic to try other paths like in the shell_tests__dir_fd().

Also add some debug messages to inform users why it's skipped.

---
 tools/perf/tests/pe-file-parsing.c | 13 +++++++++++--
 1 file changed, 11 insertions(+), 2 deletions(-)
tools/perf/tests/pe-file-parsing.c +11 -2
index 3979f38bf127c218..78dae7e43603b9f8 100644
@@ -74,10 +74,17 @@ static int test__pe_file_parsing(struct test_suite *test __maybe_unused,
 {
 	struct stat st;
 	char path_dir[PATH_MAX];
+	static const char * const tests_dirs[] = {
+		"./tools/perf/tests",
+		"./tests",
+		"./source/tests"
+	};
 
 	/* First try development tree tests. */
-	if (!lstat("./tests", &st))
-		return run_dir("./tests");
+	for (size_t i = 0; i < ARRAY_SIZE(tests_dirs); ++i) {
+		if (!lstat(tests_dirs[i], &st))
+			return run_dir(tests_dirs[i]);
+	}
 
 	/* Then installed path. */
 	snprintf(path_dir, PATH_MAX, "%s/tests", get_argv_exec_path());
@@ -85,6 +92,7 @@ static int test__pe_file_parsing(struct test_suite *test __maybe_unused,
 	if (!lstat(path_dir, &st))
 		return run_dir(path_dir);
 
+	pr_debug("cannot find the tests directory\n");
 	return TEST_SKIP;
 }
 
@@ -93,6 +101,7 @@ static int test__pe_file_parsing(struct test_suite *test __maybe_unused,
 static int test__pe_file_parsing(struct test_suite *test __maybe_unused,
 			  int subtest __maybe_unused)
 {
+	pr_debug("libbfd is not available\n");
 	return TEST_SKIP;
 }
 
Signature · 1 line
2.55.0
#8 lore raw

Re: [PATCH v2 4/4] perf test: Update PE file test for other locations

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider: - [Medium] perf/tests: spurious test failure due to brittle directory checking --

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] perf/tests: spurious test failure due to brittle directory checking --

commit 073d3f0f55cefe06ae3b0caf163e2a6505e656b3
Author: Namhyung Kim <[email protected]>

perf test: Update PE file test for other locations

This updates test__pe_file_parsing() to search multiple directories to find the tests directory, similar to shell_tests__dir_fd(), and adds debug messages when skipped.

Quoted 10 lines
tools/perf/tests/pe-file-parsing.c +2 -0
index 3979f38bf127c..78dae7e43603b 100644
@@ -74,10 +74,17 @@ static int test__pe_file_parsing(struct test_suite *test __maybe_unused,
 {
 	struct stat st;
 	char path_dir[PATH_MAX];
+	static const char * const tests_dirs[] = {
+		"./tools/perf/tests",

tools/perf/tests/pe-file-parsing.c static int test__pe_file_parsing(struct test_suite *test __maybe_unused, · line 86

Quoted 10 lines
tools/perf/tests/pe-file-parsing.c +6 -2
+		"./tests",
+		"./source/tests"
+	};
 
 	/* First try development tree tests. */
-	if (!lstat("./tests", &st))
-		return run_dir("./tests");
+	for (size_t i = 0; i < ARRAY_SIZE(tests_dirs); ++i) {
+		if (!lstat(tests_dirs[i], &st))
+			return run_dir(tests_dirs[i]);

[Severity: Medium]
Will this cause a spurious test failure if a matching directory exists but doesn't contain pe-file.exe?

The new loop eagerly calls run_dir() upon finding a matching directory name instead of verifying the actual test file exists.

If pe-file.exe is missing from the matched directory, run_dir() calls filename__read_build_id() which returns an error. This causes TEST_ASSERT_VAL to immediately return TEST_FAIL, breaking the test suite instead of falling back to the installed path.

The commit message notes this logic tries other paths like shell_tests__dir_fd(), but shell_tests__dir_fd() merely returns a directory descriptor and gracefully continues with zero shell tests if the directory is empty. By contrast, run_dir() explicitly asserts file validity and aborts the suite with a hard TEST_FAIL.

Should this verify pe-file.exe actually exists before committing to a directory and calling run_dir()?

Quoted 4 lines
tools/perf/tests/pe-file-parsing.c +1 -0
+	}
 
 	/* Then installed path. */
 	snprintf(path_dir, PATH_MAX, "%s/tests", get_argv_exec_path());
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=4
author maintainer
#9 lore raw

Re: [PATCH v2 4/4] perf test: Update PE file test for other locations

Quoted 46 lines · On Mon, Oct 12, 2026 at 05:38:20AM +0000, [email protected] wrote:

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] perf/tests: spurious test failure due to brittle directory checking --

commit 073d3f0f55cefe06ae3b0caf163e2a6505e656b3
Author: Namhyung Kim <[email protected]>

perf test: Update PE file test for other locations

This updates test__pe_file_parsing() to search multiple directories to find the tests directory, similar to shell_tests__dir_fd(), and adds debug messages when skipped.

Quoted 10 lines
tools/perf/tests/pe-file-parsing.c +2 -0
index 3979f38bf127c..78dae7e43603b 100644
@@ -74,10 +74,17 @@ static int test__pe_file_parsing(struct test_suite *test __maybe_unused,
 {
 	struct stat st;
 	char path_dir[PATH_MAX];
+	static const char * const tests_dirs[] = {
+		"./tools/perf/tests",

tools/perf/tests/pe-file-parsing.c static int test__pe_file_parsing(struct test_suite *test __maybe_unused, · line 86

Quoted 10 lines
tools/perf/tests/pe-file-parsing.c +6 -2
+		"./tests",
+		"./source/tests"
+	};
 
 	/* First try development tree tests. */
-	if (!lstat("./tests", &st))
-		return run_dir("./tests");
+	for (size_t i = 0; i < ARRAY_SIZE(tests_dirs); ++i) {
+		if (!lstat(tests_dirs[i], &st))
+			return run_dir(tests_dirs[i]);

[Severity: Medium]
Will this cause a spurious test failure if a matching directory exists but doesn't contain pe-file.exe?

The new loop eagerly calls run_dir() upon finding a matching directory name instead of verifying the actual test file exists.

If pe-file.exe is missing from the matched directory, run_dir() calls filename__read_build_id() which returns an error. This causes TEST_ASSERT_VAL to immediately return TEST_FAIL, breaking the test suite instead of falling back to the installed path.

Quoted 8 lines

The commit message notes this logic tries other paths like shell_tests__dir_fd(), but shell_tests__dir_fd() merely returns a directory descriptor and gracefully continues with zero shell tests if the directory is empty. By contrast, run_dir() explicitly asserts file validity and aborts the suite with a hard TEST_FAIL.

Should this verify pe-file.exe actually exists before committing to a directory and calling run_dir()?

It's expected to point to the tests directory in the source tree, so it would contain the file. But in a rare case where it runs on a directory which happens to have a "tests", the test will fail on reading build-id.

I think it's a minor pre-existing issue and maybe we can just ignore it.

Thanks,
Namhyung

reviewer
#10 lore raw

Re: [PATCH v2 4/4] perf test: Update PE file test for other locations

In reply to Namhyung Kim
Quoted 17 lines · Namhyung Kim

The test__pe_file_parsing() expects the test is in the tools/perf directory, but it could be root or a separate build directory. Let's update the logic to try other paths like in the shell_tests__dir_fd().

Also add some debug messages to inform users why it's skipped.

---
 tools/perf/tests/pe-file-parsing.c | 13 +++++++++++--
 1 file changed, 11 insertions(+), 2 deletions(-)
tools/perf/tests/pe-file-parsing.c +0 -0
index 3979f38bf127c218..78dae7e43603b9f8 100644

tools/perf/tests/pe-file-parsing.c static int test__pe_file_parsing(struct test_suite *test __maybe_unused, · line 80

Quoted 7 lines
tools/perf/tests/pe-file-parsing.c +4 -0
@@ -74,10 +74,17 @@ static int test__pe_file_parsing(struct test_suite *test __maybe_unused,
 {
        struct stat st;
        char path_dir[PATH_MAX];
+       static const char * const tests_dirs[] = {
+               "./tools/perf/tests",
+               "./tests",
+               "./source/tests"

I'm not sure where "source/tests" comes from?
nit: could we add the exec_path case to the list?

Thanks,
Ian

Quoted 31 lines
tools/perf/tests/pe-file-parsing.c +7 -3
+       };
 
        /* First try development tree tests. */
-       if (!lstat("./tests", &st))
-               return run_dir("./tests");
+       for (size_t i = 0; i < ARRAY_SIZE(tests_dirs); ++i) {
+               if (!lstat(tests_dirs[i], &st))
+                       return run_dir(tests_dirs[i]);
+       }
 
        /* Then installed path. */
        snprintf(path_dir, PATH_MAX, "%s/tests", get_argv_exec_path());
@@ -85,6 +92,7 @@ static int test__pe_file_parsing(struct test_suite *test __maybe_unused,
        if (!lstat(path_dir, &st))
                return run_dir(path_dir);
 
+       pr_debug("cannot find the tests directory\n");
        return TEST_SKIP;
 }
 
@@ -93,6 +101,7 @@ static int test__pe_file_parsing(struct test_suite *test __maybe_unused,
 static int test__pe_file_parsing(struct test_suite *test __maybe_unused,
                          int subtest __maybe_unused)
 {
+       pr_debug("libbfd is not available\n");
        return TEST_SKIP;
 }
 
--

2.55.0

author maintainer
#11 lore raw

Re: [PATCH v2 4/4] perf test: Update PE file test for other locations

Quoted 13 lines · Ian Rogers
Quoted 12 lines · Namhyung Kim

The test__pe_file_parsing() expects the test is in the tools/perf directory, but it could be root or a separate build directory. Let's update the logic to try other paths like in the shell_tests__dir_fd().

Also add some debug messages to inform users why it's skipped.

---
 tools/perf/tests/pe-file-parsing.c | 13 +++++++++++--
 1 file changed, 11 insertions(+), 2 deletions(-)
Quoted 14 lines

tools/perf/tests/pe-file-parsing.c static int test__pe_file_parsing(struct test_suite *test __maybe_unused, · line 80

Quoted 12 lines
tools/perf/tests/pe-file-parsing.c +4 -0
index 3979f38bf127c218..78dae7e43603b9f8 100644
@@ -74,10 +74,17 @@ static int test__pe_file_parsing(struct test_suite *test __maybe_unused,
 {
        struct stat st;
        char path_dir[PATH_MAX];
+       static const char * const tests_dirs[] = {
+               "./tools/perf/tests",
+               "./tests",
+               "./source/tests"

I'm not sure where "source/tests" comes from?

AFAIK it's set when you pass O=$(DIR) option when you build perf and it points the source tree.

Quoted 1 line

nit: could we add the exec_path case to the list?

It's checked right after. :)

Thanks,
Namhyung

Quoted 35 lines

Thanks,
Ian

Quoted 31 lines
tools/perf/tests/pe-file-parsing.c +7 -3
+       };
 
        /* First try development tree tests. */
-       if (!lstat("./tests", &st))
-               return run_dir("./tests");
+       for (size_t i = 0; i < ARRAY_SIZE(tests_dirs); ++i) {
+               if (!lstat(tests_dirs[i], &st))
+                       return run_dir(tests_dirs[i]);
+       }
 
        /* Then installed path. */
        snprintf(path_dir, PATH_MAX, "%s/tests", get_argv_exec_path());
@@ -85,6 +92,7 @@ static int test__pe_file_parsing(struct test_suite *test __maybe_unused,
        if (!lstat(path_dir, &st))
                return run_dir(path_dir);
 
+       pr_debug("cannot find the tests directory\n");
        return TEST_SKIP;
 }
 
@@ -93,6 +101,7 @@ static int test__pe_file_parsing(struct test_suite *test __maybe_unused,
 static int test__pe_file_parsing(struct test_suite *test __maybe_unused,
                          int subtest __maybe_unused)
 {
+       pr_debug("libbfd is not available\n");
        return TEST_SKIP;
 }
 
--

2.55.0

reviewer
#12 lore raw

Re: [PATCH v2 4/4] perf test: Update PE file test for other locations

Quoted 30 lines · Namhyung Kim
Quoted 13 lines · Ian Rogers
Quoted 12 lines · Namhyung Kim

The test__pe_file_parsing() expects the test is in the tools/perf directory, but it could be root or a separate build directory. Let's update the logic to try other paths like in the shell_tests__dir_fd().

Also add some debug messages to inform users why it's skipped.

---
 tools/perf/tests/pe-file-parsing.c | 13 +++++++++++--
 1 file changed, 11 insertions(+), 2 deletions(-)
Quoted 14 lines

tools/perf/tests/pe-file-parsing.c static int test__pe_file_parsing(struct test_suite *test __maybe_unused, · line 80

Quoted 12 lines
tools/perf/tests/pe-file-parsing.c +4 -0
index 3979f38bf127c218..78dae7e43603b9f8 100644
@@ -74,10 +74,17 @@ static int test__pe_file_parsing(struct test_suite *test __maybe_unused,
 {
        struct stat st;
        char path_dir[PATH_MAX];
+       static const char * const tests_dirs[] = {
+               "./tools/perf/tests",
+               "./tests",
+               "./source/tests"

I'm not sure where "source/tests" comes from?

Quoted 2 lines

AFAIK it's set when you pass O=$(DIR) option when you build perf and it points the source tree.

Ah, Andi's change:
https://lore.kernel.org/r/[email protected]

Quoted 3 lines
Quoted 1 line

nit: could we add the exec_path case to the list?

It's checked right after. :)

Right, I think you can avoid the duplicated logic by having tests_dir also have path_dir as the last entry. It'd mean always doing the snprintf but performance doesn't matter much in a test.

Thanks,
Ian

Quoted 38 lines

Thanks,
Namhyung

Quoted 35 lines

Thanks,
Ian

Quoted 31 lines
tools/perf/tests/pe-file-parsing.c +7 -3
+       };
 
        /* First try development tree tests. */
-       if (!lstat("./tests", &st))
-               return run_dir("./tests");
+       for (size_t i = 0; i < ARRAY_SIZE(tests_dirs); ++i) {
+               if (!lstat(tests_dirs[i], &st))
+                       return run_dir(tests_dirs[i]);
+       }
 
        /* Then installed path. */
        snprintf(path_dir, PATH_MAX, "%s/tests", get_argv_exec_path());
@@ -85,6 +92,7 @@ static int test__pe_file_parsing(struct test_suite *test __maybe_unused,
        if (!lstat(path_dir, &st))
                return run_dir(path_dir);
 
+       pr_debug("cannot find the tests directory\n");
        return TEST_SKIP;
 }
 
@@ -93,6 +101,7 @@ static int test__pe_file_parsing(struct test_suite *test __maybe_unused,
 static int test__pe_file_parsing(struct test_suite *test __maybe_unused,
                          int subtest __maybe_unused)
 {
+       pr_debug("libbfd is not available\n");
        return TEST_SKIP;
 }
 
--

2.55.0

author maintainer
#13 lore raw

Re: [PATCH v2 4/4] perf test: Update PE file test for other locations

Quoted 37 lines · Ian Rogers
Quoted 30 lines · Namhyung Kim
Quoted 13 lines · Ian Rogers
Quoted 12 lines · Namhyung Kim

The test__pe_file_parsing() expects the test is in the tools/perf directory, but it could be root or a separate build directory. Let's update the logic to try other paths like in the shell_tests__dir_fd().

Also add some debug messages to inform users why it's skipped.

---
 tools/perf/tests/pe-file-parsing.c | 13 +++++++++++--
 1 file changed, 11 insertions(+), 2 deletions(-)
Quoted 14 lines

tools/perf/tests/pe-file-parsing.c static int test__pe_file_parsing(struct test_suite *test __maybe_unused, · line 80

Quoted 12 lines
tools/perf/tests/pe-file-parsing.c +4 -0
index 3979f38bf127c218..78dae7e43603b9f8 100644
@@ -74,10 +74,17 @@ static int test__pe_file_parsing(struct test_suite *test __maybe_unused,
 {
        struct stat st;
        char path_dir[PATH_MAX];
+       static const char * const tests_dirs[] = {
+               "./tools/perf/tests",
+               "./tests",
+               "./source/tests"

I'm not sure where "source/tests" comes from?

Quoted 2 lines

AFAIK it's set when you pass O=$(DIR) option when you build perf and it points the source tree.

Ah, Andi's change:
https://lore.kernel.org/r/[email protected]

Quoted 7 lines
Quoted 3 lines
Quoted 1 line

nit: could we add the exec_path case to the list?

It's checked right after. :)

Right, I think you can avoid the duplicated logic by having tests_dir also have path_dir as the last entry. It'd mean always doing the snprintf but performance doesn't matter much in a test.

Ah ok, will do!

Thanks,
Namhyung

#14 lore raw
In reply to Namhyung Kim
Sashiko has reviewed this patch and found no issues. It looks great!
Quoted 10 lines

The dso__load() should return a positive number (number of symbols loaded) when it succeded to load symbols. But dso__load_bfd_symbols() only returns 0 and it's hard to check if it actually loaded symbols or not. Let's return the value of dso__has_symbol() and check it in dso__load() so that it can return 1 for success.

Fixes eac9a4342e5447cad ("perf symbols: Try reading the symbol table with libbfd")

Sashiko has reviewed this patch and found no issues. It looks great!

Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=1
reviewer Reviewed-by on v2 1/4
#15 lore raw
In reply to Namhyung Kim
Quoted 11 lines · Namhyung Kim

The dso__load() should return a positive number (number of symbols loaded) when it succeded to load symbols. But dso__load_bfd_symbols() only returns 0 and it's hard to check if it actually loaded symbols or not. Let's return the value of dso__has_symbol() and check it in dso__load() so that it can return 1 for success.

Fixes eac9a4342e5447cad ("perf symbols: Try reading the symbol table with libbfd")

Thanks,
Ian

Quoted 50 lines
---
 tools/perf/tests/pe-file-parsing.c | 2 +-
 tools/perf/util/libbfd.c           | 2 +-
 tools/perf/util/symbol.c           | 4 ++--
 3 files changed, 4 insertions(+), 4 deletions(-)
tools/perf/tests/pe-file-parsing.c +1 -1
index 30c7da79e109b480..3979f38bf127c218 100644
@@ -59,7 +59,7 @@ static int run_dir(const char *d)
        TEST_ASSERT_VAL("Failed to get dso", dso);
 
        ret = dso__load_bfd_symbols(dso, debugfile);
-       TEST_ASSERT_VAL("Failed to load symbols", ret == 0);
+       TEST_ASSERT_VAL("Failed to load symbols", ret > 0);
 
        dso__sort_by_name(dso);
        sym = dso__find_symbol_by_name(dso, "main", &idx);
tools/perf/util/libbfd.c +1 -1
index a907d7c1ac948808..a4dcb406c79e9cee 100644
@@ -463,7 +463,7 @@ int dso__load_bfd_symbols(struct dso *dso, const char *debugfile)
        symbols__fixup_duplicate(dso__symbols(dso));
        dso__set_adjust_symbols(dso, true);
 
-       err = 0;
+       err = dso__has_symbols(dso);
 out_free:
        free(symbols);
 out_close:
tools/perf/util/symbol.c +2 -3
index 5d98888d068c741a..a394e742f9b18812 100644
@@ -1930,8 +1930,8 @@ int dso__load(struct dso *dso, struct map *map)
                if (nsexit)
                        nsinfo__mountns_enter(dso__nsinfo(dso), &nsc);
 
-               if (bfdrc == 0) {
-                       ret = 0;
+               if (bfdrc > 0) {
+                       ret = 1;
                        break;
                }
 
--

2.55.0