Re: [PATCH i-g-t v4 1/4] tests/intel/xe_wedged: Ignore more dmesg warnings

[Matt Roper <matthew.d.roper@intel.com>]

On Tue, Nov 26, 2024 at 07:10:20PM +0100, Kamil Konieczny wrote:
> There are more warnings in dmesg which should be ignored,
> one comes from driver, for example:
> 
> xe 0000:00:02.0: [drm] *ERROR* GT0: reset failed (-ECANCELED)
> xe 0000:00:02.0: [drm] *ERROR* GT1: Failed to submit gsc proxy rq (-ETIME)
> 
> or from drm:
> 
> <4> [438.651441]  ? __pfx___drm_printfn_seq_file+0x10/0x10 [drm]
> <4> [438.651476]  ? __pfx___drm_puts_seq_file+0x10/0x10 [drm]
> 
> or from kernel itself:
> 
> <4> [226.597358] Modules linked in:
> 
> so extend regex with them.
> 
> v2: move '|' to begin of strings for better reading (Jonathan)
>   extend regex with two more, correct __pfx___ regex (Kamil)
>   added r-b and correcting name in Cc
> v3: fix and simplify pfx___drm regex (Kamil)
> 
> Closes: https://gitlab.freedesktop.org/drm/xe/kernel/-/issues/2919

I didn't realize you were already working on this issue; I just sent a
different fix for this issue here before I saw your patch:

https://lore.kernel.org/all/20241126230240.1780940-1-matthew.d.roper@intel.com/

Basically the test isn't touching the 'verbose' setting for fault
injection today, so we usually wind up with verbose=2 by default which
prints both the general fault injection message (not a problem) and a
kernel dump_stack (which is at warning log level and thus confuses the
runner).  If we explicitly set verbose=1, then we'll just get the
general notification about a fault being injected, but the unnecessary
dump_stack() is skipped.


Matt

> Cc: Jonathan Cavitt <jonathan.cavitt@intel.com>
> Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
> Cc: Zbigniew Kempczyński <zbigniew.kempczynski@intel.com>
> Signed-off-by: Kamil Konieczny <kamil.konieczny@linux.intel.com>
> Reviewed-by: Jonathan Cavitt <jonathan.cavitt@intel.com>
> Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
> ---
>  tests/intel/xe_wedged.c | 6 +++++-
>  1 file changed, 5 insertions(+), 1 deletion(-)
> 
> diff --git a/tests/intel/xe_wedged.c b/tests/intel/xe_wedged.c
> index 88e5d47f2..9a95a2455 100644
> --- a/tests/intel/xe_wedged.c
> +++ b/tests/intel/xe_wedged.c
> @@ -29,7 +29,11 @@
>  static void ignore_wedged_in_dmesg(void)
>  {
>  	/* this is needed for igt_runner so it will ignore it */
> -	igt_emit_ignore_dmesg_regex("CRITICAL: Xe has declared device [0-9A-Fa-f:.]* as wedged");
> +	igt_emit_ignore_dmesg_regex("CRITICAL: Xe has declared device [0-9A-Fa-f:.]* as wedged"
> +				    "|GT[0-9A-Fa-f]*: reset failed .-ECANCELED"
> +				    "|GT[0-9A-Fa-f]*: Failed to submit"
> +				    "|Modules linked in:"
> +				    "|__pfx___drm_");
>  }
>  
>  static void force_wedged(int fd)
> -- 
> 2.34.1
> 

-- 
Matt Roper
Graphics Software Engineer
Linux GPU Platform Enablement
Intel Corporation