Intel-XE Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: sashiko-bot@kernel.org
To: "Tejas Upadhyay" <tejas.upadhyay@intel.com>,
	intel-xe@lists.freedesktop.org
Subject: Re: [PATCH] drm/xe/nvl-p: Add Wa_14028564128
Date: Wed, 18 Nov 2026 15:11:42 +0000	[thread overview]
Message-ID: <sashiko-outbox-221298@kernel.org> (raw)
In-Reply-To: <20261118145429.3281059-2-tejas.upadhyay@intel.com>

Thank you for your contribution! Sashiko AI review found 2 potential
issues to consider:
- [Low] drm/xe: missing wa_14028564128_fixup_pat_value() in
  xe_pat_dump_sw_config()
- [Low] drm/xe: redundant xe_pat_wa_14028564128_safe_userptr_index()
  checks
--

commit 683d82e78ad6b286b430cc2236f6f9fa00bba693
Author: Tejas Upadhyay <tejas.upadhyay@intel.com>

drm/xe/nvl-p: Add Wa_14028564128

This commit implements Wa_14028564128 for graphics version 35.10 steppings
A0..B0 by setting XE2_NO_PROMOTE on PAT index 2 and allowing PAT index 2 in
VM_BIND and VM_MADVISE ioctl checks.

> diff --git a/drivers/gpu/drm/xe/xe_pat.c b/drivers/gpu/drm/xe/xe_pat.c
> index 5938fef22579a..49c776da4e9f3 100644
> --- a/drivers/gpu/drm/xe/xe_pat.c
> +++ b/drivers/gpu/drm/xe/xe_pat.c
[ ... ]
> @@ -420,6 +450,8 @@ static void program_pat_mcr(struct xe_gt *gt, const struct xe_pat_table_entry ta
>  		struct xe_reg_mcr reg_mcr = XE_REG_MCR(_PAT_INDEX(i));
>  		u32 pat = wa_14026539277_fixup_pat_value(gt, table[i].value);
>  
> +		pat = wa_14028564128_fixup_pat_value(gt, i, pat);
> +
>  		xe_gt_mcr_multicast_write(gt, reg_mcr, pat);
>  	}
>  

[Severity: Low]
Should xe_pat_dump_sw_config() also call wa_14028564128_fixup_pat_value()?

xe_pat_dump_sw_config() calls wa_14026539277_fixup_pat_value() and
xe_pat_wa_14026539277_reserved(), so without wa_14028564128_fixup_pat_value()
the pat_sw_config debugfs node reports PAT[2] as unreserved with
no_promote = 0, disagreeing with the hardware register value programmed by
program_pat_mcr() and dumped by xe_pat_dump().

> diff --git a/drivers/gpu/drm/xe/xe_vm.c b/drivers/gpu/drm/xe/xe_vm.c
> index dd855876f571c..80ccaa44720ce 100644
> --- a/drivers/gpu/drm/xe/xe_vm.c
> +++ b/drivers/gpu/drm/xe/xe_vm.c
> @@ -3888,7 +3888,8 @@ static int vm_bind_ioctl_check_args(struct xe_device *xe, struct xe_vm *vm,
>  		    XE_IOCTL_DBG(xe, xe_device_is_l2_flush_optimized(xe) &&
>  				 (op == DRM_XE_VM_BIND_OP_MAP_USERPTR ||
>  				  is_cpu_addr_mirror) &&
> -				 (pat_index != 19 && coh_mode != XE_COH_2WAY)) ||
> +				 (pat_index != 19 && coh_mode != XE_COH_2WAY &&
> +				  !xe_pat_wa_14028564128_safe_userptr_index(xe, pat_index))) ||
>  		    XE_IOCTL_DBG(xe, comp_en &&
>  				 op == DRM_XE_VM_BIND_OP_MAP_USERPTR) ||
>  		    XE_IOCTL_DBG(xe, op == DRM_XE_VM_BIND_OP_MAP_USERPTR &&

[Severity: Low]
Is the xe_pat_wa_14028564128_safe_userptr_index() check here (and in
xe_vm_bind_ioctl_validate_bo() and xe_vm_madvise_ioctl()) redundant?

coh_mode comes from xe_pat_index_get_coh_mode(), which returns the static
xe->pat.table[pat_index].coh_mode value.  In xe3p_lpg_pat_table, entry 2 has
coh_mode == XE_COH_2WAY, while xe_pat_wa_14028564128_safe_userptr_index()
only returns true when pat_index == 2.  When pat_index == 2,
coh_mode != XE_COH_2WAY is already false and short-circuits before calling
xe_pat_wa_14028564128_safe_userptr_index().  Exempting PAT index 2 inside
xe_pat_wa_14026539277_reserved() already allows it through unless
xe_pat_index_get_coh_mode() itself is changed to report the remapped
1-way coherency mode.

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261118145429.3281059-2-tejas.upadhyay@intel.com?part=1

      reply	other threads:[~2026-11-18 15:11 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-11-18 14:54 [PATCH] drm/xe/nvl-p: Add Wa_14028564128 Tejas Upadhyay
2026-11-18 15:11 ` sashiko-bot [this message]

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=sashiko-outbox-221298@kernel.org \
    --to=sashiko-bot@kernel.org \
    --cc=intel-xe@lists.freedesktop.org \
    --cc=sashiko-reviews@lists.linux.dev \
    --cc=tejas.upadhyay@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