Skip to contentSkip to conversation

perf test code-reading: Fix duplicate kernel maps and machine leak

PATCH Applied Ian Rogers 3 participants 3 messages Sep 28 perf events

lore
Share & tools
lore

Ian Rogers

author reviewer
#0 lore raw

PATCH 1/1 perf test code-reading: Fix duplicate kernel maps and machine leak

In do_test_code_reading(), machine__new_host(&host_env) already calls machine__create_kernel_maps(machine). Calling machine__create_kernel_maps(machine) a second time immediately afterwards duplicates all kernel module maps in machine->kmaps, causing dso__load_kcore()'s module validation (do_validate_kcore_modules()) to fail and fall back to /proc/kallsyms.

Furthermore, when try_kcore is true and !have_kcore, returning TEST_CODE_READING_NO_KCORE directly bypasses out_err, leaking the entire struct machine, its kernel/module maps and DSOs, all kallsyms symbols, and host_env.

Remove the redundant machine__create_kernel_maps() call, check for !machine after machine__new_host(), and jump to out_err when returning TEST_CODE_READING_NO_KCORE.

Fixes 0fd4008ed755 ("perf test: Use machine__new_host in mmap thread code reading test")
Assisted-by Antigravity:gemini-3.1-pro
---
 tools/perf/tests/code-reading.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)
tools/perf/tests/code-reading.c +6 -6
index f0e8ea8754ef..c50b526f44fb 100644
@@ -716,10 +716,8 @@ static int do_test_code_reading(bool try_kcore)
 
 	perf_env__init(&host_env);
 	machine = machine__new_host(&host_env);
-
-	ret = machine__create_kernel_maps(machine);
-	if (ret < 0) {
-		pr_debug("machine__create_kernel_maps failed\n");
+	if (!machine) {
+		pr_debug("machine__new_host failed\n");
 		goto out_err;
 	}
 
@@ -739,8 +737,10 @@ static int do_test_code_reading(bool try_kcore)
 	have_kcore = dso__is_kcore(dso);
 
 	/* 2nd time through we just try kcore */
-	if (try_kcore && !have_kcore)
-		return TEST_CODE_READING_NO_KCORE;
+	if (try_kcore && !have_kcore) {
+		err = TEST_CODE_READING_NO_KCORE;
+		goto out_err;
+	}
 
 	/* No point getting kernel events if there is no kernel object */
 	if (!have_vmlinux && !have_kcore)
Signature · 1 line
2.56.0.rc1.315.gc6ed9934b7-goog
#1 lore raw
Sashiko has reviewed this patch and found no issues. It looks great!
Quoted 12 lines

In do_test_code_reading(), machine__new_host(&host_env) already calls machine__create_kernel_maps(machine). Calling machine__create_kernel_maps(machine) a second time immediately afterwards duplicates all kernel module maps in machine->kmaps, causing dso__load_kcore()'s module validation (do_validate_kcore_modules()) to fail and fall back to /proc/kallsyms.

Furthermore, when try_kcore is true and !have_kcore, returning TEST_CODE_READING_NO_KCORE directly bypasses out_err, leaking the entire struct machine, its kernel/module maps and DSOs, all kallsyms symbols, and host_env.

Quoted 7 lines

Remove the redundant machine__create_kernel_maps() call, check for !machine after machine__new_host(), and jump to out_err when returning TEST_CODE_READING_NO_KCORE.

Fixes 0fd4008ed755 ("perf test: Use machine__new_host in mmap thread code reading test")
Assisted-by Antigravity:gemini-3.1-pro

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
maintainer Applied
#2 lore raw
In reply to Ian Rogers
Quoted 11 lines · Ian Rogers

Furthermore, when try_kcore is true and !have_kcore, returning TEST_CODE_READING_NO_KCORE directly bypasses out_err, leaking the entire struct machine, its kernel/module maps and DSOs, all kallsyms symbols, and host_env.

Remove the redundant machine__create_kernel_maps() call, check for !machine after machine__new_host(), and jump to out_err when returning TEST_CODE_READING_NO_KCORE.

Fixes 0fd4008ed755 ("perf test: Use machine__new_host in mmap thread code reading test")
Assisted-by Antigravity:gemini-3.1-pro

Thanks, applied to perf-tools-next, for v7.4.

- Arnaldo