lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20250827-imperial-amigurumi-malkoha-b99a9d@houat>
Date: Wed, 27 Aug 2025 11:52:06 +0200
From: Maxime Ripard <mripard@...nel.org>
To: Jocelyn Falempe <jfalempe@...hat.com>
Cc: Maarten Lankhorst <maarten.lankhorst@...ux.intel.com>, 
	Thomas Zimmermann <tzimmermann@...e.de>, David Airlie <airlied@...il.com>, 
	Simona Vetter <simona@...ll.ch>, Javier Martinez Canillas <javierm@...hat.com>, 
	linux-kernel@...r.kernel.org, dri-devel@...ts.freedesktop.org
Subject: Re: [PATCH 3/3] drm/panic: Add a kconfig option to dump kunits
 results to png

Hi,

On Thu, Aug 21, 2025 at 11:49:07AM +0200, Jocelyn Falempe wrote:
> This is a bit hacky, but very handy if you want to customize the
> panic screen.
> It allows to dump the generated images to the logs, and then a python
> script can convert it to .png files. It makes it easy to check how
> the panic screen will look on different resolutions, without having
> to crash a VM.
> To not pollute the logs, it uses a monochrome framebuffer, compress
> it with zlib, and base64 encode it.
> 
> Signed-off-by: Jocelyn Falempe <jfalempe@...hat.com>
> ---
>  drivers/gpu/drm/Kconfig.debug          |  14 ++++
>  drivers/gpu/drm/tests/drm_panic_test.c | 111 +++++++++++++++++++++++++
>  scripts/kunitpanic2png.py              |  53 ++++++++++++
>  3 files changed, 178 insertions(+)
>  create mode 100755 scripts/kunitpanic2png.py
> 
> diff --git a/drivers/gpu/drm/Kconfig.debug b/drivers/gpu/drm/Kconfig.debug
> index 05dc43c0b8c5..d8ae85132d32 100644
> --- a/drivers/gpu/drm/Kconfig.debug
> +++ b/drivers/gpu/drm/Kconfig.debug
> @@ -84,6 +84,20 @@ config DRM_KUNIT_TEST
>  
>  	  If in doubt, say "N".
>  
> +config DRM_PANIC_KUNIT_TEST_DUMP
> +	bool "Enable screen dump to logs in KUnit tests for drm_panic"
> +	default n
> +	depends on DRM && DRM_PANIC && DRM_KUNIT_TEST
> +	select ZLIB_DEFLATE
> +	help
> +	  This allows to dump the panic screen to the KUnit tests logs.
> +	  It's possible with a small python script to write pngs from the logs.
> +
> +	  This is only to help developers customizing the drm_panic screen,
> +	  checking the result for different resolutions.
> +
> +	  If in doubt, say "N"
> +
>  config DRM_TTM_KUNIT_TEST
>  	tristate "KUnit tests for TTM" if !KUNIT_ALL_TESTS
>  	default n
> diff --git a/drivers/gpu/drm/tests/drm_panic_test.c b/drivers/gpu/drm/tests/drm_panic_test.c
> index 46ff3e5e0e5d..8cddb845aea9 100644
> --- a/drivers/gpu/drm/tests/drm_panic_test.c
> +++ b/drivers/gpu/drm/tests/drm_panic_test.c
> @@ -115,24 +115,135 @@ static void drm_test_panic_screen_user_page(struct kunit *test)
>  	kfree(pages);
>  }
>  
> +#ifdef CONFIG_DRM_PANIC_KUNIT_TEST_DUMP
> +#include <linux/base64.h>
> +#include <linux/delay.h>
> +#include <linux/zlib.h>
> +
> +#define LINE_LEN 128
> +
> +#define COMPR_LEVEL 6
> +#define WINDOW_BITS 12
> +#define MEM_LEVEL 4
> +
> +static int compress_image(u8 *src, int size, u8 *dst)
> +{
> +	struct z_stream_s stream;
> +
> +	stream.workspace = kmalloc(zlib_deflate_workspacesize(WINDOW_BITS, MEM_LEVEL),
> +				   GFP_KERNEL);
> +
> +	if (zlib_deflateInit2(&stream, COMPR_LEVEL, Z_DEFLATED, WINDOW_BITS,
> +			      MEM_LEVEL, Z_DEFAULT_STRATEGY) != Z_OK)
> +		return -EINVAL;
> +
> +	stream.next_in = src;
> +	stream.avail_in = size;
> +	stream.total_in = 0;
> +	stream.next_out = dst;
> +	stream.avail_out = size;
> +	stream.total_out = 0;
> +
> +	if (zlib_deflate(&stream, Z_FINISH) != Z_STREAM_END)
> +		return -EINVAL;
> +
> +	if (zlib_deflateEnd(&stream) != Z_OK)
> +		return -EINVAL;
> +
> +	kfree(stream.workspace);
> +
> +	return stream.total_out;
> +}
> +
> +static void dump_image(u8 *fb, unsigned int width, unsigned int height)
> +{
> +	int len = 0;
> +	char *dst;
> +	char *compressed;
> +	int sent = 0;
> +	int stride = DIV_ROUND_UP(width, 8);
> +	int size = stride * height;
> +
> +	compressed = vzalloc(size);
> +	if (!compressed)
> +		return;
> +	len = compress_image(fb, size, compressed);
> +	if (len < 0) {
> +		pr_err("Compression failed %d", len);
> +		return;
> +	}
> +
> +	dst = vzalloc(4 * DIV_ROUND_UP(len, 3) + 1);
> +	if (!dst)
> +		return;
> +
> +	len = base64_encode(compressed, len, dst);
> +
> +	pr_info("KUNIT PANIC IMAGE DUMP START %dx%d", width, height);
> +	while (len > 0) {
> +		char save = dst[sent + LINE_LEN];
> +
> +		dst[sent + LINE_LEN] = 0;
> +		pr_info("%s", dst + sent);
> +		dst[sent + LINE_LEN] = save;
> +		sent += LINE_LEN;
> +		len -= LINE_LEN;
> +	}
> +	pr_info("KUNIT PANIC IMAGE DUMP END");

The kunit test output format is defined, and we should probably use a
diagnostic line for this:
https://docs.kernel.org/dev-tools/ktap.html#diagnostic-lines

We should probably cc the kunit maintainers about this too.

Maxime

Download attachment "signature.asc" of type "application/pgp-signature" (274 bytes)

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ