Skip to contentSkip to conversation

bpf: Refactoring btf_name_valid_identifier() and btf_name_valid_section()

PATCH In review Jeongjun Park 2 participants 2 messages Aug 19, 2024 – Aug 21, 2024 BPF

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 bpf: Refactoring btf_name_valid_identifier() and btf_name_valid_section()

Currently, btf_name_valid_identifier() and btf_name_valid_section() are written in a while loop and use pointer operations, so it takes a long time to understand the operation of the code. Therefore, I suggest refactoring the code to make it easier to maintain.

In addition, btf_name_valid_section() does not check for the case where src[0] is a NULL value, resulting in an out-of-bounds vuln. Therefore, a check for this should be added.

Fixes bd70a8fb7ca4 ("bpf: Allow all printable characters in BTF DATASEC names")
---
 kernel/bpf/btf.c | 27 ++++++++++++---------------
 1 file changed, 12 insertions(+), 15 deletions(-)
kernel/bpf/btf.c +12 -15
index 674b38c33c74..c1e2aead9141 100644
@@ -794,21 +794,18 @@ static bool btf_name_valid_identifier(const struct btf *btf, u32 offset)
 {
 	/* offset must be valid */
 	const char *src = btf_str_by_offset(btf, offset);
-	const char *src_limit;
+	int i;
 
-	if (!__btf_name_char_ok(*src, true))
+	if (!__btf_name_char_ok(src[0], true))
 		return false;
 
 	/* set a limit on identifier length */
-	src_limit = src + KSYM_NAME_LEN;
-	src++;
-	while (*src && src < src_limit) {
-		if (!__btf_name_char_ok(*src, false))
+	for (i = 1; i < KSYM_NAME_LEN && src[i]; i++) {
+		if (!__btf_name_char_ok(src[i], false))
 			return false;
-		src++;
 	}
 
-	return !*src;
+	return !src[i];
 }
 
 /* Allow any printable character in DATASEC names */
@@ -816,18 +813,18 @@ static bool btf_name_valid_section(const struct btf *btf, u32 offset)
 {
 	/* offset must be valid */
 	const char *src = btf_str_by_offset(btf, offset);
-	const char *src_limit;
+	int i;
+
+	if (!src[0])
+		return false;
 
 	/* set a limit on identifier length */
-	src_limit = src + KSYM_NAME_LEN;
-	src++;
-	while (*src && src < src_limit) {
-		if (!isprint(*src))
+	for (i = 1; i < KSYM_NAME_LEN && src[i]; i++) {
+		if (!isprint(src[i]))
 			return false;
-		src++;
 	}
 
-	return !*src;
+	return !src[i];
 }
 
 static const char *__btf_name_by_offset(const struct btf *btf, u32 offset)

--

#1 lore raw
Quoted 5 lines · Jeongjun Park

Currently, btf_name_valid_identifier() and btf_name_valid_section() are written in a while loop and use pointer operations, so it takes a long time to understand the operation of the code. Therefore, I suggest refactoring the code to make it easier to maintain.

imo it's harder to read after refactoring. Pls avoid.

Quoted 3 lines

In addition, btf_name_valid_section() does not check for the case where src[0] is a NULL value, resulting in an out-of-bounds vuln. Therefore, a check for this should be added.

Hmm. Not sure about it. Pls add a selftest that demonstrates the issue and produce a patch to fix just that.
Do not mix it with questionable refactoring.

pw-bot: cr