Re: [PATCH 1/1] drm/xe: Add support for WA 14028250015

[Ceraolo Spurio, Daniele]

On 11/3/2026 6:33 AM, Gustavo Sousa wrote:
"Teres Alexis, Alan Previn" <alan.previn.teres.ale...@intel.com> writes:

[snip]

Also, on the subject, I would prefer that we stick to the notation
Wa_14028250015.
will fix
[snip]

        GUC_WA_KLV_REMAP_RANGED_TLB_INV                                         
        = 0x900f,
        GUC_WA_KLV_IGNORE_MMIO_READ_SEM_TOKEN_64                                
        = 0x9010,
        GUC_WA_KLV_IGNORE_MULTIQ_CTX_COMPLETE_WITH_PREEMPT_IN_CSB               
        = 0x9011,
+       GUC_WA_KLV_ENABLE_PAGE_RECLAIM_ERROR_COMPLETION                         
        = 0x9012,
Probably better to use the same name as defined by GuC here?
if u mean the internal spec - sure i can do that but i didnt realize that is 
the convention here.
I'm not sure if this is the convention though. :-)

I just thought it would make sense to keep the names in sync.  Maybe
Daniele can provide his input here.

We do try to match what's in the GuC specs when we can, but we allow 
divergence if it makes sense. In this case I think using 
GUC_WA_KLV_PAGE_RECLAMATION_COMPLETION_TIMEOUT to match the GuC specs is 
the better option.

Daniele


--
Gustavo Sousa

[snip]

+       /* On GuC 70.78 onwards, we enable GuC PRL-completion-IRQ workaround on 
applicable GMIDs */
+       if (XE_GT_WA(gt, 14028250015) && GUC_FIRMWARE_VER_AT_LEAST(&gt->uc.guc, 70, 78) 
&&
+           gt_to_xe(gt)->info.has_page_reclaim_hw_assist)
Should we really gate this by the has_page_reclaim_hw_assist feature
flag?  This flag can be changed at runtime via debugfs.  I would assume
that GuC would do the right thing if PRL is not supported on the IP it
is running.
yeah - like your idea better - will just always do it.

[snip]