[DTrace-devel] [PATCH] The fbtprovider/tst.return0.d test should not use ioctl

[Eugene Loh]

Reviewed-by: Eugene Loh <eugene.loh at oracle.com>
But is there a wider problem here we want to address?  In the -Z patch, 
triggers were introduced for a number of tests that were waiting on 
ioctl.  Maybe that's fine?

On 6/22/21 1:26 AM, Kris Van Hees wrote:
> This test was triggering on ioctl because that was the system call used
> by DTrace to read trace data.  It was therefore guaranteed to be called
> frequently.
>
> With the new implementation ioctl is no longer used.  Instead, we can
> trigger on the poll system call.
>
> Signed-off-by: Kris Van Hees<kris.van.hees at oracle.com>
> ---
>   test/unittest/fbtprovider/tst.return0.d | 4 +---
>   1 file changed, 1 insertion(+), 3 deletions(-)
>
> diff --git a/test/unittest/fbtprovider/tst.return0.d b/test/unittest/fbtprovider/tst.return0.d
> index bcfde573..7d1c7491 100644
> --- a/test/unittest/fbtprovider/tst.return0.d
> +++ b/test/unittest/fbtprovider/tst.return0.d
> @@ -11,14 +11,12 @@
>    * SECTION: FBT Provider/Probe arguments
>    */
>   
> -/* @@xfail: dtv2 */
>   /* @@runtest-opts: -Z */
>   
>   #pragma D option quiet
>   #pragma D option statusrate=10ms
>   
> -fbt::SyS_ioctl:return,
> -fbt::__x64_sys_ioctl:return
> +fbt::do_sys_poll:return
>   /arg1 == 0/
>   {
>   	printf("%s %x returned 0", probefunc, arg0);
> -- 2.31.1