Skip to contentSkip to conversation

perf trace: Add upper bound checks for augmented BTF struct printing

PATCH Applied Aaron Tomlin 3 participants 4 messages Sep 6 – Sep 8 perf events

v1 of 3 Applied Latest is v3 →

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 perf trace: Add upper bound checks for augmented BTF struct printing

When pretty-printing augmented struct payloads using BTF via btf_struct_scnprintf(), augmented_arg->size is currently only validated against values <= 0.

However, several edge cases can result in size mismatches or buffer over-reads:
1. If arg->augmented.size is smaller than sizeof(*augmented_arg), dereferencing augmented_arg->size reads past the available buffer.

2. If augmented_arg->size exceeds arg->augmented.size - sizeof(*augmented_arg), calculating consumed = sizeof(*augmented_arg) + augmented_arg->size can overflow signed integer limits (e.g., with crafted INT_MAX values in an untrusted perf.data file) or cause arg->augmented.size to underflow. This advances arg->augmented.args out of bounds, corrupting the parsing state for subsequent arguments in multi-argument syscalls.

3. If the captured payload is truncated (e.g., short reads in BPF, or during cross-architecture analysis such as replaying a 32-bit perf.data on a 64-bit host where host BTF type->size exceeds the 32-bit target payload), passing type->size to btf_dump__dump_type_data() causes libbpf to read past the end of the payload buffer.

Enforce an upper bound on augmented_arg->size against the remaining buffer (arg->augmented.size - sizeof(*augmented_arg)) and verify that the captured payload contains at least type->size bytes before passing it to btf_dump__dump_type_data().

Fixes cb32035214b9 ("perf trace: Pretty print struct data")
---
 tools/perf/builtin-trace.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)
tools/perf/builtin-trace.c +3 -2
index be19d70eba09..20fffc24507b 100644
@@ -1133,12 +1133,13 @@ static size_t btf_struct_scnprintf(const struct btf_type *type, struct btf *btf,
 	LIBBPF_OPTS(btf_dump_opts, dump_opts);
 	LIBBPF_OPTS(btf_dump_type_data_opts, dump_data_opts);
 
-	if (arg == NULL || arg->augmented.args == NULL || arg->augmented.size <= 0 ||
+	if (arg == NULL || arg->augmented.args == NULL || arg->augmented.size < (int)sizeof(*augmented_arg) ||
 	    arg->fmt == NULL || !arg->fmt->from_user)
 		return 0;
 
 	augmented_arg = arg->augmented.args;
-	if (augmented_arg->size <= 0)
+	if (augmented_arg->size <= 0 || augmented_arg->size > arg->augmented.size - (int)sizeof(*augmented_arg) ||
+	    (size_t)augmented_arg->size < type->size)
 		return 0;
 
 	dump_data_opts.compact	  = true;

base-commit: 02f6847e1822714a4201b87e42f92b0d43e8549d

Signature · 1 line
2.55.0
#1 lore raw
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:

Pre-existing issues:
- [High] The patch adds bounds checking to btf_struct_scnprintf to prevent buffer over-reads and integer overflows from untrusted perf.data files, but identical vulnerabilities remain unfixed in several other augmented argument formatters. --

commit 292e31352e85e8bac1a8cec2bc218077f4783d42
Author: Aaron Tomlin <[email protected]>

perf trace: Add upper bound checks for augmented BTF struct printing

This commit adds bounds checking to btf_struct_scnprintf() when pretty-printing augmented struct payloads. It prevents buffer over-reads and integer overflow when parsing untrusted perf.data files.

Quoted 8 lines
tools/perf/builtin-trace.c +0 -0
index be19d70eba092..20fffc24507b5 100644
@@ -1133,12 +1133,13 @@ static size_t btf_struct_scnprintf(const struct btf_type *type, struct btf *btf,
 	LIBBPF_OPTS(btf_dump_opts, dump_opts);
 	LIBBPF_OPTS(btf_dump_type_data_opts, dump_data_opts);
 

tools/perf/builtin-trace.c static size_t btf_struct_scnprintf(const struct btf_type *type, struct btf *btf, · line 1143

Quoted 10 lines
tools/perf/builtin-trace.c +3 -2
-	if (arg == NULL || arg->augmented.args == NULL || arg->augmented.size <= 0 ||
+	if (arg == NULL || arg->augmented.args == NULL || arg->augmented.size < (int)sizeof(*augmented_arg) ||
 	    arg->fmt == NULL || !arg->fmt->from_user)
 		return 0;
 
 	augmented_arg = arg->augmented.args;
-	if (augmented_arg->size <= 0)
+	if (augmented_arg->size <= 0 || augmented_arg->size > arg->augmented.size - (int)sizeof(*augmented_arg) ||
+	    (size_t)augmented_arg->size < type->size)
 		return 0;

[Severity: High]
This is a pre-existing issue, but while this change correctly adds bounds checking to prevent buffer over-reads and integer overflows in btf_struct_scnprintf(), identical vulnerabilities remain unfixed in several other augmented argument formatters.

When a crafted perf.data file provides an artificially large augmented_arg->size or a truncated payload buffer, could the following formatters still be vulnerable to out-of-bounds reads or state corruption?

In tools/perf/builtin-trace.c:
syscall_arg__scnprintf_augmented_string():

    size_t printed = scnprintf(bf, size, "\"%.*s\"", augmented_arg->size,
                               augmented_arg->value);
    int consumed = sizeof(*augmented_arg) + augmented_arg->size;

This lacks validation that arg->augmented.size is >= sizeof(*augmented_arg), fails to bound augmented_arg->size against the remaining payload buffer, and performs a pointer advance calculation that can overflow and corrupt parsing state.

In tools/perf/builtin-trace.c:
syscall_arg__scnprintf_buf():

    for (int j = 0; j < augmented_arg->size; ++j) {

This unconditionally iterates up to augmented_arg->size without bounds checking against the remaining payload buffer arg->augmented.size, leading to buffer over-reads.

In tools/perf/trace/beauty/perf_event_open.c:
syscall_arg__scnprintf_augmented_perf_event_attr():

    struct perf_event_attr *attr = (void *)arg->augmented.args->value;

This fails to ensure arg->augmented.size is large enough to contain the perf_event_attr struct before casting and reading from it.

In tools/perf/trace/beauty/sockaddr.c:
syscall_arg__scnprintf_augmented_sockaddr():

    struct sockaddr *sa = (struct sockaddr *)&arg->augmented.args->value;

This reads fields like sa->sa_family without any payload size validation, leading to potential out-of-bounds reads.

In tools/perf/trace/beauty/timespec.c:
syscall_arg__scnprintf_augmented_timespec():

    struct timespec *ts = (struct timespec *)arg->augmented.args->value;

This dereferences struct fields without verifying that the payload contains at least sizeof(struct timespec) bytes.

Would it make sense to update these other formatters to include the same bounds checking being added here?

Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=1
#2 lore raw
Quoted 79 lines · On Sun, Sep 06, 2026 at 01:23:58AM +0000, [email protected] wrote:

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:

Pre-existing issues:
- [High] The patch adds bounds checking to btf_struct_scnprintf to prevent buffer over-reads and integer overflows from untrusted perf.data files, but identical vulnerabilities remain unfixed in several other augmented argument formatters. --

commit 292e31352e85e8bac1a8cec2bc218077f4783d42
Author: Aaron Tomlin <[email protected]>

perf trace: Add upper bound checks for augmented BTF struct printing

This commit adds bounds checking to btf_struct_scnprintf() when pretty-printing augmented struct payloads. It prevents buffer over-reads and integer overflow when parsing untrusted perf.data files.

Quoted 8 lines
tools/perf/builtin-trace.c +0 -0
index be19d70eba092..20fffc24507b5 100644
@@ -1133,12 +1133,13 @@ static size_t btf_struct_scnprintf(const struct btf_type *type, struct btf *btf,
 	LIBBPF_OPTS(btf_dump_opts, dump_opts);
 	LIBBPF_OPTS(btf_dump_type_data_opts, dump_data_opts);
 

tools/perf/builtin-trace.c static size_t btf_struct_scnprintf(const struct btf_type *type, struct btf *btf, · line 1143

Quoted 10 lines
tools/perf/builtin-trace.c +3 -2
-	if (arg == NULL || arg->augmented.args == NULL || arg->augmented.size <= 0 ||
+	if (arg == NULL || arg->augmented.args == NULL || arg->augmented.size < (int)sizeof(*augmented_arg) ||
 	    arg->fmt == NULL || !arg->fmt->from_user)
 		return 0;
 
 	augmented_arg = arg->augmented.args;
-	if (augmented_arg->size <= 0)
+	if (augmented_arg->size <= 0 || augmented_arg->size > arg->augmented.size - (int)sizeof(*augmented_arg) ||
+	    (size_t)augmented_arg->size < type->size)
 		return 0;

[Severity: High]
This is a pre-existing issue, but while this change correctly adds bounds checking to prevent buffer over-reads and integer overflows in btf_struct_scnprintf(), identical vulnerabilities remain unfixed in several other augmented argument formatters.

When a crafted perf.data file provides an artificially large augmented_arg->size or a truncated payload buffer, could the following formatters still be vulnerable to out-of-bounds reads or state corruption?

In tools/perf/builtin-trace.c:
syscall_arg__scnprintf_augmented_string():

    size_t printed = scnprintf(bf, size, "\"%.*s\"", augmented_arg->size,
                               augmented_arg->value);
    int consumed = sizeof(*augmented_arg) + augmented_arg->size;

This lacks validation that arg->augmented.size is >= sizeof(*augmented_arg), fails to bound augmented_arg->size against the remaining payload buffer, and performs a pointer advance calculation that can overflow and corrupt parsing state.

In tools/perf/builtin-trace.c:
syscall_arg__scnprintf_buf():

    for (int j = 0; j < augmented_arg->size; ++j) {

This unconditionally iterates up to augmented_arg->size without bounds checking against the remaining payload buffer arg->augmented.size, leading to buffer over-reads.

In tools/perf/trace/beauty/perf_event_open.c:
syscall_arg__scnprintf_augmented_perf_event_attr():

    struct perf_event_attr *attr = (void *)arg->augmented.args->value;

This fails to ensure arg->augmented.size is large enough to contain the perf_event_attr struct before casting and reading from it.

In tools/perf/trace/beauty/sockaddr.c:
syscall_arg__scnprintf_augmented_sockaddr():

    struct sockaddr *sa = (struct sockaddr *)&arg->augmented.args->value;

This reads fields like sa->sa_family without any payload size validation, leading to potential out-of-bounds reads.

In tools/perf/trace/beauty/timespec.c:
syscall_arg__scnprintf_augmented_timespec():

Quoted 10 lines
    struct timespec *ts = (struct timespec *)arg->augmented.args->value;

This dereferences struct fields without verifying that the payload contains at least sizeof(struct timespec) bytes.

Would it make sense to update these other formatters to include the same bounds checking being added here?

Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=1

Acknowledged.

Kind regards,

Signature · 1 line
Aaron Tomlin
maintainer Applied
#3 lore raw
Quoted 85 lines · Aaron Tomlin
Quoted 84 lines · On Sun, Sep 06, 2026 at 01:23:58AM +0000, [email protected] wrote:

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:

Pre-existing issues:
- [High] The patch adds bounds checking to btf_struct_scnprintf to prevent buffer over-reads and integer overflows from untrusted perf.data files, but identical vulnerabilities remain unfixed in several other augmented argument formatters. --

commit 292e31352e85e8bac1a8cec2bc218077f4783d42
Author: Aaron Tomlin <[email protected]>

perf trace: Add upper bound checks for augmented BTF struct printing

This commit adds bounds checking to btf_struct_scnprintf() when pretty-printing augmented struct payloads. It prevents buffer over-reads and integer overflow when parsing untrusted perf.data files.

Quoted 8 lines
tools/perf/builtin-trace.c +0 -0
index be19d70eba092..20fffc24507b5 100644
@@ -1133,12 +1133,13 @@ static size_t btf_struct_scnprintf(const struct btf_type *type, struct btf *btf,
 	LIBBPF_OPTS(btf_dump_opts, dump_opts);
 	LIBBPF_OPTS(btf_dump_type_data_opts, dump_data_opts);
 

tools/perf/builtin-trace.c static size_t btf_struct_scnprintf(const struct btf_type *type, struct btf *btf, · line 1143

Quoted 10 lines
tools/perf/builtin-trace.c +3 -2
-	if (arg == NULL || arg->augmented.args == NULL || arg->augmented.size <= 0 ||
+	if (arg == NULL || arg->augmented.args == NULL || arg->augmented.size < (int)sizeof(*augmented_arg) ||
 	    arg->fmt == NULL || !arg->fmt->from_user)
 		return 0;
 
 	augmented_arg = arg->augmented.args;
-	if (augmented_arg->size <= 0)
+	if (augmented_arg->size <= 0 || augmented_arg->size > arg->augmented.size - (int)sizeof(*augmented_arg) ||
+	    (size_t)augmented_arg->size < type->size)
 		return 0;

[Severity: High]
This is a pre-existing issue, but while this change correctly adds bounds checking to prevent buffer over-reads and integer overflows in btf_struct_scnprintf(), identical vulnerabilities remain unfixed in several other augmented argument formatters.

When a crafted perf.data file provides an artificially large augmented_arg->size or a truncated payload buffer, could the following formatters still be vulnerable to out-of-bounds reads or state corruption?

In tools/perf/builtin-trace.c:
syscall_arg__scnprintf_augmented_string():

    size_t printed = scnprintf(bf, size, "\"%.*s\"", augmented_arg->size,
                               augmented_arg->value);
    int consumed = sizeof(*augmented_arg) + augmented_arg->size;

This lacks validation that arg->augmented.size is >= sizeof(*augmented_arg), fails to bound augmented_arg->size against the remaining payload buffer, and performs a pointer advance calculation that can overflow and corrupt parsing state.

In tools/perf/builtin-trace.c:
syscall_arg__scnprintf_buf():

    for (int j = 0; j < augmented_arg->size; ++j) {

This unconditionally iterates up to augmented_arg->size without bounds checking against the remaining payload buffer arg->augmented.size, leading to buffer over-reads.

In tools/perf/trace/beauty/perf_event_open.c:
syscall_arg__scnprintf_augmented_perf_event_attr():

    struct perf_event_attr *attr = (void *)arg->augmented.args->value;

This fails to ensure arg->augmented.size is large enough to contain the perf_event_attr struct before casting and reading from it.

In tools/perf/trace/beauty/sockaddr.c:
syscall_arg__scnprintf_augmented_sockaddr():

    struct sockaddr *sa = (struct sockaddr *)&arg->augmented.args->value;

This reads fields like sa->sa_family without any payload size validation, leading to potential out-of-bounds reads.

In tools/perf/trace/beauty/timespec.c:
syscall_arg__scnprintf_augmented_timespec():

    struct timespec *ts = (struct timespec *)arg->augmented.args->value;

This dereferences struct fields without verifying that the payload contains at least sizeof(struct timespec) bytes.

Quoted 7 lines
Quoted 5 lines

Would it make sense to update these other formatters to include the same bounds checking being added here?

Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=1

Acknowledged.

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

- Arnaldo