[PATCH] selftests: tls: fix read/write of garbage value

[PATCH] selftests: tls: fix read/write of garbage value

[Ankit Khushwaha]

In 'poll_partial_rec_async' a uninitialized char variable 'token' with
*garbage* value is used for write/read instruction to synchronize
between threads via a pipe.

tls.c:2833:26: warning: variable 'token' is uninitialized
      when passed as a const pointer argument here
      [-Wuninitialized-const-pointer]
 2833 |            EXPECT_EQ(write(p[1], &token, 1), 1); /* Barrier #1 */

Initialize 'token' to '\0' to prevent write/read of garbage value.

Signed-off-by: Ankit Khushwaha <ankitkhushwaha.linux@gmail.com>
---
compiler used: clang version 21.1.5 (Fedora 21.1.5-1.fc43).
---
 tools/testing/selftests/net/tls.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tools/testing/selftests/net/tls.c b/tools/testing/selftests/net/tls.c
index a3ef4b57eb5f..a4d16a460fbe 100644
--- a/tools/testing/selftests/net/tls.c
+++ b/tools/testing/selftests/net/tls.c
@@ -2786,10 +2786,10 @@ TEST_F(tls_err, epoll_partial_rec)
 TEST_F(tls_err, poll_partial_rec_async)
 {
 	struct pollfd pfd = { };
+	char token = '\0';
 	ssize_t rec_len;
 	char rec[256];
 	char buf[128];
-	char token;
 	int p[2];
 	int ret;

--
2.52.0

Re: [PATCH] selftests: tls: fix read/write of garbage value

[Sabrina Dubroca]

2025-11-29, 12:07:26 +0530, Ankit Khushwaha wrote:
> In 'poll_partial_rec_async' a uninitialized char variable 'token' with
> *garbage* value is used for write/read instruction to synchronize
> between threads via a pipe.
> 
> tls.c:2833:26: warning: variable 'token' is uninitialized
>       when passed as a const pointer argument here
>       [-Wuninitialized-const-pointer]
>  2833 |            EXPECT_EQ(write(p[1], &token, 1), 1); /* Barrier #1 */
> 
> Initialize 'token' to '\0' to prevent write/read of garbage value.

I'm not opposed to making the compiler happy, but in this case
"garbage" is completely fine, we don't care about the value.

So I think the subject and the commit message should talk about
"silencing a compiler warning" rather than "fixing use of garbage
value".

And your patch should indicate the target tree in the subject (here,
with [PATCH net-next]), see
https://www.kernel.org/doc/html/next/process/maintainer-netdev.html

-- 
Sabrina

Re: [PATCH] selftests: tls: fix read/write of garbage value

[Ankit Khushwaha]

On Sat, Nov 29, 2025 at 11:22:06AM +0100, Sabrina Dubroca wrote:
> I'm not opposed to making the compiler happy, but in this case
> "garbage" is completely fine, we don't care about the value.
> 
> So I think the subject and the commit message should talk about
> "silencing a compiler warning" rather than "fixing use of garbage
> value".
 
> And your patch should indicate the target tree in the subject (here,
> with [PATCH net-next]), see
> https://www.kernel.org/doc/html/next/process/maintainer-netdev.html

Hi Sabrina,
thanks for the review, i will send a v2 patch with suggested changes.

--
Ankit