[PATCH v2] resolv: Disable TCP servers in tst-resolv-res_init-failure

[Carlos O'Donell]

On 7/21/26 2:50 AM, Florian Weimer wrote:
> This avoids spurious xaccept errors on slow systems.  With this change,
> the test still reproduces the original bug.

LGTM.

Reviewed-by: Carlos O'Donell <carlos@redhat.com>
  
> ---
> v2: Commented updated.
>   resolv/tst-resolv-res_init-failure.c | 17 ++++++++++++++++-
>   1 file changed, 16 insertions(+), 1 deletion(-)
> 
> diff --git a/resolv/tst-resolv-res_init-failure.c b/resolv/tst-resolv-res_init-failure.c
> index d20348acf2..b3cb0b70fe 100644
> --- a/resolv/tst-resolv-res_init-failure.c
> +++ b/resolv/tst-resolv-res_init-failure.c
> @@ -66,7 +66,22 @@ static int
>   do_test (void)
>   {
>     struct resolv_test *aux = resolv_test_start
> -    ((struct resolv_redirect_config) { .response_callback = response });
> +    ((struct resolv_redirect_config)
> +     {
> +       .response_callback = response,
> +       /* Disable TCP servers to prevent test failures due to a race
> +	  condition: If the initial setrlimit call below executes
> +	  before the kernel has allocated the client FD in the accept
> +	  system call on the TCP DNS server threads, the system call
> +	  fails even if no client connections arrive.  (The FD is
> +	  pre-allocated.)  */

OK. That is enough to highlight the conflict is with the xaccept in the
server thread.

> +       .servers =
> +	 {
> +	   { .disable_tcp = true, },
> +	   { .disable_tcp = true, },
> +	   { .disable_tcp = true, },
> +	 },
> +     });
>   
>     /* Initial lookup.  This drives __res_context_send through its init
>        block, which allocates _u._ext.nsaddrs[] and sets
> 
> base-commit: 249639137ed360d1827d9e663261ba26d2b01d81
>