[PATCH i-g-t v3 2/3] tests/kms_atomic_transition: Reducing debug loglevel dynamically

[Modem, Bhanuprakash]

On 11-07-2024 09:43 am, Pranay Samala wrote:
> This test is debug logs are too much and was killed
> due to exceeding disk usage limit on the less disk
> space machines.
> 
> So dynamically reducing the debug log level to 13.
> 
> v3:
> - Calling loglevel init functionality in lib file (Bhanu)
> - Here 2 means the log level will be reduced by 2 (Bhanu)

Please keep maintain the rev history (v2 is missing).

> 
> Signed-off-by: Pranay Samala <pranay.samala at intel.com>
> ---
>   tests/kms_atomic_transition.c | 3 +++
>   1 file changed, 3 insertions(+)
> 
> diff --git a/tests/kms_atomic_transition.c b/tests/kms_atomic_transition.c
> index 29dd8ac4e..f89cd0814 100644
> --- a/tests/kms_atomic_transition.c
> +++ b/tests/kms_atomic_transition.c
> @@ -34,6 +34,7 @@
>   #include "igt_rand.h"
>   #include "drmtest.h"
>   #include "sw_sync.h"
> +#include "igt_sysfs.h"
>   #include <errno.h>
>   #include <pthread.h>
>   #include <stdbool.h>
> @@ -1176,6 +1177,8 @@ igt_main_args("", long_opts, help_str, opt_handler, &data)
>   	igt_fixture {
>   		data.drm_fd = drm_open_driver_master(DRIVER_ANY);
>   
> +		igt_drm_debug_level_update(2);

Comments in Rev 2 is not addressed.

Probably, you can read the current debug level & take the decision.

Ex:
if (igt_drm_debug_level_get() > <your magic number>)
	igt_drm_debug_level_update(2);

Also, please move this logic to end of the igt_fixture.
Ex: If there is valid connector found, then there is no point to update 
the debug level.

- Bhanu

> +
>   		kmstest_set_vt_graphics_mode();
>   
>   		igt_display_require(&data.display, data.drm_fd);