Intel-XE Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
To: Matthew Auld <matthew.auld@intel.com>,
	Maarten Lankhorst <dev@lankhorst.se>,
	<intel-xe@lists.freedesktop.org>
Subject: Re: [PATCH] drm/xe: Only probe stolen memory when display support is enabled.
Date: Thu, 27 Feb 2025 09:04:39 -0800	[thread overview]
Message-ID: <28c3265b-3baf-49d8-b5f3-47a49b4972d3@intel.com> (raw)
In-Reply-To: <b10d1bfe-60dd-4412-ba72-7af698e593b0@intel.com>



On 2/27/2025 7:05 AM, Matthew Auld wrote:
> On 27/02/2025 14:55, Maarten Lankhorst wrote:
>> Without display, there is no need for display stolen memory.
>>
>> Signed-off-by: Maarten Lankhorst <dev@lankhorst.se>
>
> There is also xe_gsc_init_post_hwconfig() which looks to use stolen, 
> and I don't think that is dependent on display. But maybe that gsc 
> stuff doesn't have to use stolen? Daniele, any idea here?

The GSC allocation does have to use stolen, at least on some platforms. 
I think the requirement might be gone from recent platforms (PTL 
onwards) because the FW is now being copied to WOPCM instead, but I am 
not 100% sure. I didn't confirm this because we still need to keep using 
stolen for older platforms and it was just easier to keep using stolen 
everywhere; if you think there is value in changing this for newer 
platforms I can check with the GSC team and confirm.

Daniele

>
>> ---
>>   drivers/gpu/drm/xe/xe_ttm_stolen_mgr.c | 2 +-
>>   1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/drivers/gpu/drm/xe/xe_ttm_stolen_mgr.c 
>> b/drivers/gpu/drm/xe/xe_ttm_stolen_mgr.c
>> index d9c9d2547aadf..2e60d5346ec43 100644
>> --- a/drivers/gpu/drm/xe/xe_ttm_stolen_mgr.c
>> +++ b/drivers/gpu/drm/xe/xe_ttm_stolen_mgr.c
>> @@ -218,7 +218,7 @@ int xe_ttm_stolen_mgr_init(struct xe_device *xe)
>>       if (!mgr)
>>           return -ENOMEM;
>>   -    if (IS_SRIOV_VF(xe))
>> +    if (!xe->info.probe_display)
>>           stolen_size = 0;
>>       else if (IS_DGFX(xe))
>>           stolen_size = detect_bar2_dgfx(xe, mgr);
>


  reply	other threads:[~2025-02-27 17:05 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-02-27 14:55 [PATCH] drm/xe: Only probe stolen memory when display support is enabled Maarten Lankhorst
2025-02-27 15:05 ` Matthew Auld
2025-02-27 17:04   ` Daniele Ceraolo Spurio [this message]
2025-02-27 15:31 ` ✓ CI.Patch_applied: success for " Patchwork
2025-02-27 15:31 ` ✓ CI.checkpatch: " Patchwork
2025-02-27 15:32 ` ✓ CI.KUnit: " Patchwork
2025-02-27 15:49 ` ✓ CI.Build: " Patchwork
2025-02-27 15:52 ` ✓ CI.Hooks: " Patchwork
2025-02-27 15:53 ` ✓ CI.checksparse: " Patchwork
2025-02-27 16:11 ` ✓ Xe.CI.BAT: " Patchwork
2025-02-27 19:36 ` ✗ Xe.CI.Full: failure " Patchwork

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=28c3265b-3baf-49d8-b5f3-47a49b4972d3@intel.com \
    --to=daniele.ceraolospurio@intel.com \
    --cc=dev@lankhorst.se \
    --cc=intel-xe@lists.freedesktop.org \
    --cc=matthew.auld@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