public inbox for intel-gfx@lists.freedesktop.org
 help / color / mirror / Atom feed
From: Nirmoy Das <nirmoy.das@linux.intel.com>
To: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>,
	Nirmoy Das <nirmoy.das@intel.com>,
	intel-gfx@lists.freedesktop.org
Cc: Andrzej Hajda <andrzej.hajda@intel.com>
Subject: Re: [Intel-gfx] [PATCH 2/2] drm/i915/display: Do not use stolen on MTL
Date: Wed, 12 Jul 2023 14:43:35 +0200	[thread overview]
Message-ID: <497c9b56-3315-16d8-8594-a050550647bc@linux.intel.com> (raw)
In-Reply-To: <a443b2f7-8a9b-1b8e-c6ef-61dce338ef87@linux.intel.com>


On 7/11/2023 5:59 PM, Tvrtko Ursulin wrote:
>
> On 10/07/2023 10:00, Nirmoy Das wrote:
>> Hi Tvrkto,
>>
>> On 7/6/2023 3:43 PM, Tvrtko Ursulin wrote:
>>>
>>> On 06/07/2023 14:35, Nirmoy Das wrote:
>>>>
>>>> On 7/6/2023 3:32 PM, Tvrtko Ursulin wrote:
>>>>>
>>>>> On 30/06/2023 18:01, Nirmoy Das wrote:
>>>>>> Use smem on MTL due to a HW bug in MTL that prevents
>>>>>> reading from stolen memory using LMEM BAR.
>>>>>
>>>>> Does anything remain in stolen or could the memory region just not 
>>>>> be created?
>>>>
>>>>
>>>> GSC requires DSM which can't use smem for another bug.
>>>
>>> Okay, thanks.
>>>
>>> As a related comment, these if-if-if object creation ladders were 
>>> always a bit ugly and some years ago I was suggesting we create a 
>>> helper with some "intent/usage" flags. Which could then dtrt ie. 
>>> create the right object for that intent/usage and platform. I 
>>> *think* I possibly even had a RFC... need to try and find it.
>>
>>
>> Did you find it :) Would be nice to have a better way to detect and 
>> apply memory region as per platfrom/usecase.
>
> Nope. Basically the idea boiled down to figuring out if it is possible 
> to express the "requirements" via intent flags. Like do we need CPU 
> access, is it mostly GPU, can it be volatile etc. And then combine the 
> intent with the platform to figure out what kind of object to create. 


Yes, that would be nice.

> But it was many years ago and I am not sure if the idea would still 
> apply so easily, without looking at the all call sites. Could easily 
> end up complicated so I cannot dare to say it is worth spending time 
> looking at this.

I will record this in a Jira and hopefully get back to this when I have 
less urgent work.


Thanks,

Nirmoy

>
> Regards,
>
> Tvrtko
>
>>
>>
>> Regards,
>>
>> Nirmoy
>>
>>>
>>> Regards,
>>>
>>> Tvrtko
>>>
>>>>
>>>> Regards,
>>>>
>>>> Nirmoy
>>>>
>>>>
>>>>>
>>>>> Regards,
>>>>>
>>>>> Tvrtko
>>>>>
>>>>>> Cc: Oak Zeng <oak.zeng@intel.com>
>>>>>> Cc: Jani Nikula <jani.nikula@linux.intel.com>
>>>>>> Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
>>>>>> Cc: Andi Shyti <andi.shyti@linux.intel.com>
>>>>>> Cc: Andrzej Hajda <andrzej.hajda@intel.com>
>>>>>> Signed-off-by: Nirmoy Das <nirmoy.das@intel.com>
>>>>>> ---
>>>>>>   drivers/gpu/drm/i915/display/intel_fbdev.c   | 2 ++
>>>>>>   drivers/gpu/drm/i915/display/intel_overlay.c | 7 ++++---
>>>>>>   2 files changed, 6 insertions(+), 3 deletions(-)
>>>>>>
>>>>>> diff --git a/drivers/gpu/drm/i915/display/intel_fbdev.c 
>>>>>> b/drivers/gpu/drm/i915/display/intel_fbdev.c
>>>>>> index 1cc0ddc6a310..10e38d60f9ef 100644
>>>>>> --- a/drivers/gpu/drm/i915/display/intel_fbdev.c
>>>>>> +++ b/drivers/gpu/drm/i915/display/intel_fbdev.c
>>>>>> @@ -182,6 +182,8 @@ static int intelfb_alloc(struct drm_fb_helper 
>>>>>> *helper,
>>>>>>           obj = i915_gem_object_create_lmem(dev_priv, size,
>>>>>>                             I915_BO_ALLOC_CONTIGUOUS |
>>>>>>                             I915_BO_ALLOC_USER);
>>>>>> +    } else if (IS_METEORLAKE(dev_priv)) { /* Wa_22018444074 */
>>>>>> +        obj = i915_gem_object_create_shmem(dev_priv, size);
>>>>>>       } else {
>>>>>>           /*
>>>>>>            * If the FB is too big, just don't use it since fbdev 
>>>>>> is not very
>>>>>> diff --git a/drivers/gpu/drm/i915/display/intel_overlay.c 
>>>>>> b/drivers/gpu/drm/i915/display/intel_overlay.c
>>>>>> index d6fe2bbabe55..05ae446c8a56 100644
>>>>>> --- a/drivers/gpu/drm/i915/display/intel_overlay.c
>>>>>> +++ b/drivers/gpu/drm/i915/display/intel_overlay.c
>>>>>> @@ -1348,12 +1348,13 @@ int intel_overlay_attrs_ioctl(struct 
>>>>>> drm_device *dev, void *data,
>>>>>>   static int get_registers(struct intel_overlay *overlay, bool 
>>>>>> use_phys)
>>>>>>   {
>>>>>>       struct drm_i915_private *i915 = overlay->i915;
>>>>>> -    struct drm_i915_gem_object *obj;
>>>>>> +    struct drm_i915_gem_object *obj = NULL;
>>>>>>       struct i915_vma *vma;
>>>>>>       int err;
>>>>>>   -    obj = i915_gem_object_create_stolen(i915, PAGE_SIZE);
>>>>>> -    if (IS_ERR(obj))
>>>>>> +    if (!IS_METEORLAKE(i915)) /* Wa_22018444074 */
>>>>>> +        obj = i915_gem_object_create_stolen(i915, PAGE_SIZE);
>>>>>> +    if (IS_ERR_OR_NULL(obj))
>>>>>>           obj = i915_gem_object_create_internal(i915, PAGE_SIZE);
>>>>>>       if (IS_ERR(obj))
>>>>>>           return PTR_ERR(obj);

  reply	other threads:[~2023-07-12 12:43 UTC|newest]

Thread overview: 20+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-06-30 17:01 [Intel-gfx] [PATCH 1/2] drm/i915/gt: Do not use stolen on MTL Nirmoy Das
2023-06-30 17:01 ` [Intel-gfx] [PATCH 2/2] drm/i915/display: " Nirmoy Das
2023-07-05 10:02   ` Andrzej Hajda
2023-07-11 13:44     ` Nirmoy Das
2023-07-06 13:32   ` Tvrtko Ursulin
2023-07-06 13:35     ` Nirmoy Das
2023-07-06 13:43       ` Tvrtko Ursulin
2023-07-10  9:00         ` Nirmoy Das
2023-07-11 15:59           ` Tvrtko Ursulin
2023-07-12 12:43             ` Nirmoy Das [this message]
2023-07-11 14:50   ` Andi Shyti
2023-06-30 17:31 ` [Intel-gfx] ✗ Fi.CI.SPARSE: warning for series starting with [1/2] drm/i915/gt: " Patchwork
2023-06-30 17:45 ` [Intel-gfx] ✓ Fi.CI.BAT: success " Patchwork
2023-07-01  6:45 ` [Intel-gfx] ✓ Fi.CI.IGT: " Patchwork
2023-07-04 15:01 ` [Intel-gfx] [PATCH 1/2] " Zeng, Oak
2023-07-04 15:31   ` Nirmoy Das
2023-07-04 15:41     ` Zeng, Oak
2023-07-11 14:50   ` Andi Shyti
2023-07-05  9:51 ` Andrzej Hajda
2023-07-11 14:48 ` Andi Shyti

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=497c9b56-3315-16d8-8594-a050550647bc@linux.intel.com \
    --to=nirmoy.das@linux.intel.com \
    --cc=andrzej.hajda@intel.com \
    --cc=intel-gfx@lists.freedesktop.org \
    --cc=nirmoy.das@intel.com \
    --cc=tvrtko.ursulin@linux.intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox