Re: [PATCH bpf] bpf: allow UTF-8 literals in bpf_bprintf_prepare()

[Alan Maguire <alan.maguire@oracle.com>]

On 14/04/2026 16:26, Paul Chaignon wrote:
> On Tue, Apr 14, 2026 at 09:40:01AM +0800, Yihan Ding wrote:
>> bpf_bprintf_prepare() currently rejects any non-ASCII byte in the format
>> string, such as UTF-8 Chinese text.
>>
>> All BPF formatted output helpers that go through bpf_bprintf_prepare(),
>> such as bpf_trace_printk(), bpf_seq_printf() and bpf_snprintf(), only
>> need ASCII parsing for conversion specifiers. Plain text does not need
>> that restriction, but today any byte >= 0x80 makes the format fail
>> validation.
>>
>> As a result, UTF-8 text literals are rejected even when they are not part
>> of a format specifier. In practice, an ASCII-only bpf_trace_printk()
>> format works, while the same format with UTF-8 literal text produces no
>> trace output.
>>
>> Allow non-ASCII bytes in plain text while keeping the existing control
>> character checks and keeping format specifiers ASCII-only. This preserves
>> the current parsing rules for '%' sequences and allows valid UTF-8 text
>> to be emitted.
> 
> Nice! I can finally print proper French from BPF :)
> 
>>
>> Extend the trace_printk selftest accordingly by emitting both ASCII and
>> UTF-8 strings and verifying that both appear in the trace output.
>>
>> Fixes: 48cac3f4a96d ("bpf: Implement formatted output helpers with bstr_printf")
>> Signed-off-by: Yihan Ding <dingyihan@uniontech.com>
> 
> Not sure if this should go to bpf or bpf-next. I'll let the maintainers
> decide.
> 
>> ---
>> Testing:
>> - Reproduced on x86_64 without this patch: ASCII trace output works, while
>>    UTF-8 literal text in bpf_trace_printk() is rejected and produces no trace
>>    output.
>> - Verified with tools/testing/selftests/bpf: ./test_progs -t trace_printk
>>
>>  kernel/bpf/helpers.c                          | 21 +++++++++++++-----
>>  .../selftests/bpf/prog_tests/trace_printk.c   | 22 ++++++++++++++-----
>>  .../selftests/bpf/progs/trace_printk.c        |  5 +++++
>>  3 files changed, 38 insertions(+), 10 deletions(-)
>>
>> diff --git a/kernel/bpf/helpers.c b/kernel/bpf/helpers.c
>> index 6eb6c82ed2ee..e2f103297e4a 100644
>> --- a/kernel/bpf/helpers.c
>> +++ b/kernel/bpf/helpers.c
>> @@ -845,7 +845,13 @@ int bpf_bprintf_prepare(const char *fmt, u32 fmt_size, const u64 *raw_args,
>>  		data->buf = buffers->buf;
>>  
>>  	for (i = 0; i < fmt_size; i++) {
>> -		if ((!isprint(fmt[i]) && !isspace(fmt[i])) || !isascii(fmt[i])) {
>> +		unsigned char c = fmt[i];
>> +
>> +		/*
>> +		 * Permit non-ASCII bytes in plain text so UTF-8 messages can be
>> +		 * emitted, while keeping format specifiers ASCII-only.
>> +		 */
>> +		if (isascii(c) && !isprint(c) && !isspace(c)) {
>>  			err = -EINVAL;
>>  			goto out;
>>  		}
>> @@ -867,6 +873,10 @@ int bpf_bprintf_prepare(const char *fmt, u32 fmt_size, const u64 *raw_args,
>>  		 * always access fmt[i + 1], in the worst case it will be a 0
>>  		 */
>>  		i++;
>> +		if (!isascii((unsigned char)fmt[i])) {
>> +			err = -EINVAL;
>> +			goto out;
>> +		}
>>  
>>  		/* skip optional "[0 +-][num]" width formatting field */
>>  		while (fmt[i] == '0' || fmt[i] == '+'  || fmt[i] == '-' ||
>> @@ -881,8 +891,9 @@ int bpf_bprintf_prepare(const char *fmt, u32 fmt_size, const u64 *raw_args,
>>  		if (fmt[i] == 'p') {
>>  			sizeof_cur_arg = sizeof(long);
>>  
>> -			if (fmt[i + 1] == 0 || isspace(fmt[i + 1]) ||
>> -			    ispunct(fmt[i + 1])) {
>> +			if (fmt[i + 1] == 0 ||
>> +			    isspace((unsigned char)fmt[i + 1]) ||
>> +			    ispunct((unsigned char)fmt[i + 1])) {
> 
> Why is this change needed? isspace and ispunct already cast to unsigned
> char.
> 
>>  				if (tmp_buf)
>>  					cur_arg = raw_args[num_spec];
>>  				goto nocopy_fmt;
>> @@ -958,8 +969,8 @@ int bpf_bprintf_prepare(const char *fmt, u32 fmt_size, const u64 *raw_args,
>>  			fmt_ptype = fmt[i];
>>  fmt_str:
>>  			if (fmt[i + 1] != 0 &&
>> -			    !isspace(fmt[i + 1]) &&
>> -			    !ispunct(fmt[i + 1])) {
>> +			    !isspace((unsigned char)fmt[i + 1]) &&
>> +			    !ispunct((unsigned char)fmt[i + 1])) {
> 
> Same here.
> 
>>  				err = -EINVAL;
>>  				goto out;
>>  			}
>> diff --git a/tools/testing/selftests/bpf/prog_tests/trace_printk.c b/tools/testing/selftests/bpf/prog_tests/trace_printk.c
>> index e56e88596d64..f7b03dc4eaf4 100644
>> --- a/tools/testing/selftests/bpf/prog_tests/trace_printk.c
>> +++ b/tools/testing/selftests/bpf/prog_tests/trace_printk.c
>> @@ -6,18 +6,21 @@
>>  #include "trace_printk.lskel.h"
>>  
>>  #define SEARCHMSG	"testing,testing"
>> +#define SEARCHMSG_UTF8	"中文,测试"
>>  
>>  static void trace_pipe_cb(const char *str, void *data)
>>  {
>>  	if (strstr(str, SEARCHMSG) != NULL)
>> -		(*(int *)data)++;
>> +		((int *)data)[0]++;
>> +	if (strstr(str, SEARCHMSG_UTF8) != NULL)
>> +		((int *)data)[1]++;
>>  }
>>  
>>  void serial_test_trace_printk(void)
>>  {
>>  	struct trace_printk_lskel__bss *bss;
>>  	struct trace_printk_lskel *skel;
>> -	int err = 0, found = 0;
>> +	int err = 0, found[2] = {};
>>  
>>  	skel = trace_printk_lskel__open();
>>  	if (!ASSERT_OK_PTR(skel, "trace_printk__open"))
>> @@ -46,11 +49,20 @@ void serial_test_trace_printk(void)
>>  	if (!ASSERT_GT(bss->trace_printk_ret, 0, "bss->trace_printk_ret"))
>>  		goto cleanup;
>>  
>> -	/* verify our search string is in the trace buffer */
>> -	ASSERT_OK(read_trace_pipe_iter(trace_pipe_cb, &found, 1000),
>> +	if (!ASSERT_GT(bss->trace_printk_utf8_ran, 0, "bss->trace_printk_utf8_ran"))
>> +		goto cleanup;
>> +
>> +	if (!ASSERT_GT(bss->trace_printk_utf8_ret, 0, "bss->trace_printk_utf8_ret"))
>> +		goto cleanup;
>> +
>> +	/* verify our search strings are in the trace buffer */
>> +	ASSERT_OK(read_trace_pipe_iter(trace_pipe_cb, found, 1000),
>>  		 "read_trace_pipe_iter");
>>  
>> -	if (!ASSERT_EQ(found, bss->trace_printk_ran, "found"))
>> +	if (!ASSERT_EQ(found[0], bss->trace_printk_ran, "found"))
>> +		goto cleanup;
>> +
>> +	if (!ASSERT_EQ(found[1], bss->trace_printk_utf8_ran, "found_utf8"))
>>  		goto cleanup;
>>  
>>  cleanup:
>> diff --git a/tools/testing/selftests/bpf/progs/trace_printk.c b/tools/testing/selftests/bpf/progs/trace_printk.c
>> index 6695478c2b25..97afe8b149b0 100644
>> --- a/tools/testing/selftests/bpf/progs/trace_printk.c
>> +++ b/tools/testing/selftests/bpf/progs/trace_printk.c
>> @@ -10,13 +10,18 @@ char _license[] SEC("license") = "GPL";
>>  
>>  int trace_printk_ret = 0;
>>  int trace_printk_ran = 0;
>> +int trace_printk_utf8_ret = 0;
>> +int trace_printk_utf8_ran = 0;
>>  
>> -const char fmt[] = "Testing,testing %d\n";
>> +static const char fmt[] = "Testing,testing %d\n";
>> +static const char utf8_fmt[] = "中文,测试 %d\n";
> 
> The build is failing because you made these static.
> 
>>  
>>  SEC("fentry/" SYS_PREFIX "sys_nanosleep")
>>  int sys_enter(void *ctx)
>>  {
>>  	trace_printk_ret = bpf_trace_printk(fmt, sizeof(fmt),
>>  					    ++trace_printk_ran);
>> +	trace_printk_utf8_ret = bpf_trace_printk(utf8_fmt, sizeof(utf8_fmt),
>> +						 ++trace_printk_utf8_ran);
>>  	return 0;
>>  }
> > Please put the selftest coverage extensions in a second patch.
>

+1; might also be worth adding a negative test that ensures failure when
using a UTF-8 character as format specifier.
 
> pw-bot: cr
> 
>> -- 
>> 2.20.1
>>
>