Re: [PATCH 6/8] drm/vram-helper: don't use ttm bo->offset v3

From: Thomas Zimmermann <tzimmermann@suse.de>
To: Gerd Hoffmann <kraxel@redhat.com>, Nirmoy Das <nirmoy.aiemd@gmail.com>
Cc: thellstrom@vmware.com, airlied@linux.ie, kenny.ho@amd.com,
	brian.welty@intel.com, dri-devel@lists.freedesktop.org,
	nirmoy.das@amd.com, linux-graphics-maintainer@vmware.com,
	amd-gfx@lists.freedesktop.org,
	Daniel Vetter <daniel.vetter@ffwll.ch>,
	alexander.deucher@amd.com, sean@poorly.run,
	christian.koenig@amd.com, bskeggs@redhat.com
Subject: Re: [PATCH 6/8] drm/vram-helper: don't use ttm bo->offset v3
Date: Thu, 5 Mar 2020 15:26:05 +0100	[thread overview]
Message-ID: <6d7146b9-4ac1-e861-334f-950de790321d@suse.de> (raw)
In-Reply-To: <20200305140749.5oggp4gh6hozsh7j@sirius.home.kraxel.org>

[-- Attachment #1.1.1: Type: text/plain, Size: 2241 bytes --]

Hi

Am 05.03.20 um 15:07 schrieb Gerd Hoffmann:
> On Thu, Mar 05, 2020 at 02:29:08PM +0100, Nirmoy Das wrote:
>> Calculate GEM VRAM bo's offset within vram-helper without depending on
>> bo->offset.
>>
>> Signed-off-by: Nirmoy Das <nirmoy.das@amd.com>
>> Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
>> ---
>>  drivers/gpu/drm/drm_gem_vram_helper.c | 9 ++++++++-
>>  1 file changed, 8 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/gpu/drm/drm_gem_vram_helper.c b/drivers/gpu/drm/drm_gem_vram_helper.c
>> index 92a11bb42365..2749c2d25ac4 100644
>> --- a/drivers/gpu/drm/drm_gem_vram_helper.c
>> +++ b/drivers/gpu/drm/drm_gem_vram_helper.c
>> @@ -198,6 +198,13 @@ u64 drm_gem_vram_mmap_offset(struct drm_gem_vram_object *gbo)
>>  }
>>  EXPORT_SYMBOL(drm_gem_vram_mmap_offset);
>>
>> +static s64 drm_gem_vram_pg_offset(struct drm_gem_vram_object *gbo)
>> +{
>> +	if (WARN_ON_ONCE(!gbo->bo.mem.mm_node))
>> +		return 0;
> 
> returns 0 on error.
> 
>> +	return gbo->bo.mem.start;
>> +}
>> +
>>  /**
>>   * drm_gem_vram_offset() - \
>>  	Returns a GEM VRAM object's offset in video memory
>> @@ -214,7 +221,7 @@ s64 drm_gem_vram_offset(struct drm_gem_vram_object *gbo)
>>  {
>>  	if (WARN_ON_ONCE(!gbo->pin_count))
>>  		return (s64)-ENODEV;
> 
> returns -errno on error.
> 
>> -	return gbo->bo.offset;
>> +	return drm_gem_vram_pg_offset(gbo) << PAGE_SHIFT;
> 
> And given that one calls the other behavior on error should better be
> consistent ...

It is expected that the offset is valid if pin_count is positive.
Anything else would be a massive ref-counting bug. And that's been the
behavior of the old code as well.

But I agree that the current patch is inconsistent. I suggest changing
the return type of drm_gem_vram_pg_offset() to u64.

Best regards
Thomas

> 
> cheers,
>   Gerd
> 
> _______________________________________________
> dri-devel mailing list
> dri-devel@lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/dri-devel
> 

-- 
Thomas Zimmermann
Graphics Driver Developer
SUSE Software Solutions Germany GmbH
Maxfeldstr. 5, 90409 Nürnberg, Germany
(HRB 36809, AG Nürnberg)
Geschäftsführer: Felix Imendörffer

[-- Attachment #1.2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

[-- Attachment #2: Type: text/plain, Size: 160 bytes --]

_______________________________________________
dri-devel mailing list
dri-devel@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/dri-devel