linux-kernel - Re: [PATCH v2 59/59] drm: RFC - make drm_dyndbg_user.o for drm-*

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: <76b58aa3-0401-41a3-930f-c49880fefbfd@bootlin.com>
Date: Mon, 24 Mar 2025 16:00:11 +0100
From: Louis Chauvet <louis.chauvet@...tlin.com>
To: Jim Cromie <jim.cromie@...il.com>, linux-kernel@...r.kernel.org,
 dri-devel@...ts.freedesktop.org, amd-gfx@...ts.freedesktop.org,
 intel-gvt-dev@...ts.freedesktop.org, intel-gfx@...ts.freedesktop.org,
 intel-gfx-trybot@...ts.freedesktop.org
Cc: jbaron@...mai.com, gregkh@...uxfoundation.org, ukaszb@...omium.org,
 daniel.vetter@...ll.ch, tvrtko.ursulin@...ux.intel.com,
 jani.nikula@...el.com, ville.syrjala@...ux.intel.com
Subject: Re: [PATCH v2 59/59] drm: RFC - make drm_dyndbg_user.o for
 drm-*_helpers, drivers



Le 20/03/2025 à 19:52, Jim Cromie a écrit :
> Add new drm_dyndbg_user.c with a single call to
> DYNDBG_CLASSMAP_USE(drm_debug_classes).  This creates a _class_user
> record (and a linkage dependency).

I agree, this could be a very nice thing to automagically have the _USE 
call included. But if this remains a per-module work, I think it is 
better in the C code, it is less error-prone.

Maybe to clarify this is a per-module call, you can move it near the 
MODULE_* macros.

> If a driver adds this object to its Makefile target, it gets the
> record, which authorizes dyndbg to enable the module's class'd
> pr_debugs, such as DRMs <category>_dbg() macros.
> 
> So Id like to automatically inject this object into drivers.  I tried
> subdir-objs-y, but thats not a thing.
> 
> In drm/Makefile:
> 
> Add object dependency to drm_*_helper-y targets:
> 
>    $targ-$(CONFIG_DRM_USE_DYNAMIC_DEBUG) += drm_dyndbg_user.o
> 
> Attempt a foreach:
> 
>    to add $driver-y += ../drm_dyndbg_user.o
> 
> this appears to be a train-wreck for impl reasons, but it describes a
> want/need reasonably well.  It might not be a good maintainble idea.
> 
> Explicitly adding to radeon/Makefile worked:
> 
>    $radeon-$(CONFIG_DRM_USE_DYNAMIC_DEBUG) += ../drm_dyndbg_user.o
> 
> But doing this is just as per-module as just adding the _USE()
> explicitly to the main .c file, which is less magical than make-fu.
> 
> Also, it appears to cause make && make rebuilds.
> 
> and try to link it to helpers and everything
> 
> Signed-off-by: Jim Cromie <jim.cromie@...il.com>
> ---
>   drivers/gpu/drm/Makefile          |  9 +++++++++
>   drivers/gpu/drm/drm_dyndbg_user.c | 11 +++++++++++
>   2 files changed, 20 insertions(+)
>   create mode 100644 drivers/gpu/drm/drm_dyndbg_user.c
> 
> diff --git a/drivers/gpu/drm/Makefile b/drivers/gpu/drm/Makefile
> index 704f94efc804..1adb5a262180 100644
> --- a/drivers/gpu/drm/Makefile
> +++ b/drivers/gpu/drm/Makefile
> @@ -74,10 +74,12 @@ drm-y := \
>   	drm_vblank_work.o \
>   	drm_vma_manager.o \
>   	drm_writeback.o
> +
>   drm-$(CONFIG_DRM_CLIENT) += \
>   	drm_client.o \
>   	drm_client_event.o \
>   	drm_client_modeset.o
> +
>   drm-$(CONFIG_DRM_LIB_RANDOM) += lib/drm_random.o
>   drm-$(CONFIG_COMPAT) += drm_ioc32.o
>   drm-$(CONFIG_DRM_PANEL) += drm_panel.o
> @@ -111,20 +113,25 @@ obj-$(CONFIG_DRM_BUDDY) += drm_buddy.o
>   drm_dma_helper-y := drm_gem_dma_helper.o
>   drm_dma_helper-$(CONFIG_DRM_FBDEV_EMULATION) += drm_fbdev_dma.o
>   drm_dma_helper-$(CONFIG_DRM_KMS_HELPER) += drm_fb_dma_helper.o
> +drm_dma_helper-$(CONFIG_DRM_USE_DYNAMIC_DEBUG) += drm_dyndbg_user.o
>   obj-$(CONFIG_DRM_GEM_DMA_HELPER) += drm_dma_helper.o
>   
>   drm_shmem_helper-y := drm_gem_shmem_helper.o
>   drm_shmem_helper-$(CONFIG_DRM_FBDEV_EMULATION) += drm_fbdev_shmem.o
> +drm_shmem_helper-$(CONFIG_DRM_USE_DYNAMIC_DEBUG) += drm_dyndbg_user.o
>   obj-$(CONFIG_DRM_GEM_SHMEM_HELPER) += drm_shmem_helper.o
>   
>   drm_suballoc_helper-y := drm_suballoc.o
> +drm_suballoc_helper-$(CONFIG_DRM_USE_DYNAMIC_DEBUG) += drm_dyndbg_user.o
>   obj-$(CONFIG_DRM_SUBALLOC_HELPER) += drm_suballoc_helper.o
>   
>   drm_vram_helper-y := drm_gem_vram_helper.o
> +drm_vram_helper-$(CONFIG_DRM_USE_DYNAMIC_DEBUG) += drm_dyndbg_user.o
>   obj-$(CONFIG_DRM_VRAM_HELPER) += drm_vram_helper.o
>   
>   drm_ttm_helper-y := drm_gem_ttm_helper.o
>   drm_ttm_helper-$(CONFIG_DRM_FBDEV_EMULATION) += drm_fbdev_ttm.o
> +drm_ttm_helper-$(CONFIG_DRM_USE_DYNAMIC_DEBUG) += drm_dyndbg_user.o
>   obj-$(CONFIG_DRM_TTM_HELPER) += drm_ttm_helper.o
>   
>   #
> @@ -149,11 +156,13 @@ drm_kms_helper-y := \
>   	drm_simple_kms_helper.o
>   drm_kms_helper-$(CONFIG_DRM_PANEL_BRIDGE) += bridge/panel.o
>   drm_kms_helper-$(CONFIG_DRM_FBDEV_EMULATION) += drm_fb_helper.o
> +drm_kms_helper-$(CONFIG_DRM_USE_DYNAMIC_DEBUG) += drm_dyndbg_user.o
>   obj-$(CONFIG_DRM_KMS_HELPER) += drm_kms_helper.o
>   
>   #
>   # Drivers and the rest
>   #
> +subdir-obj-$(CONFIG_DRM_USE_DYNAMIC_DEBUG) += drm_dyndbg_user.o
>   
>   obj-y			+= tests/
>   
> diff --git a/drivers/gpu/drm/drm_dyndbg_user.c b/drivers/gpu/drm/drm_dyndbg_user.c
> new file mode 100644
> index 000000000000..9e4aa87d4b58
> --- /dev/null
> +++ b/drivers/gpu/drm/drm_dyndbg_user.c
> @@ -0,0 +1,11 @@
> +// SPDX-License-Identifier: GPL-2.0-or-later
> +
> +#include "drm/drm_print.h"
> +/*
> + * if DRM_USE_DYNAMIC_DEBBUG:
> + *    DYNDBG_CLASSMAP_USE(drm_debug_classes);
> + *
> + * dyndbg classmaps are opt-in, so modules which call drm:_*_dbg must
> + * link this to authorize dyndbg to change the static-keys underneath.
> + */
> +DRM_CLASSMAP_USE(drm_debug_classes);

-- 
Louis Chauvet, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com