Intel-XE Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Matthew Brost <matthew.brost@intel.com>
To: Raag Jadav <raag.jadav@intel.com>
Cc: <intel-xe@lists.freedesktop.org>,
	<daniele.ceraolospurio@intel.com>, <michal.wajdeczko@intel.com>
Subject: Re: [PATCH v1] drm/xe/guc_ct: Assert on credits mismatch during runtime suspend
Date: Tue, 2 Dec 2025 08:01:43 -0800	[thread overview]
Message-ID: <aS8NZz36I8vOpWHn@lstrano-desk.jf.intel.com> (raw)
In-Reply-To: <20251202083334.554045-1-raag.jadav@intel.com>

On Tue, Dec 02, 2025 at 02:03:34PM +0530, Raag Jadav wrote:
> G2H credits should be in fully idle state when runtime suspending GuC CT.
> Assert on mismatch.
> 
> Suggested-by: Matthew Brost <matthew.brost@intel.com>

Reviewed-by: Matthew Brost <matthew.brost@intel.com>

> Signed-off-by: Raag Jadav <raag.jadav@intel.com>
> ---
>  drivers/gpu/drm/xe/xe_guc_ct.c | 15 ++++++++++++---
>  1 file changed, 12 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/gpu/drm/xe/xe_guc_ct.c b/drivers/gpu/drm/xe/xe_guc_ct.c
> index 2a24d3fb6881..648f0f523abb 100644
> --- a/drivers/gpu/drm/xe/xe_guc_ct.c
> +++ b/drivers/gpu/drm/xe/xe_guc_ct.c
> @@ -253,8 +253,11 @@ static bool g2h_fence_needs_alloc(struct g2h_fence *g2h_fence)
>  #define CTB_DESC_SIZE		ALIGN(sizeof(struct guc_ct_buffer_desc), SZ_2K)
>  #define CTB_H2G_BUFFER_OFFSET	(CTB_DESC_SIZE * 2)
>  #define CTB_H2G_BUFFER_SIZE	(SZ_4K)
> +#define CTB_H2G_BUFFER_DWORDS	(CTB_H2G_BUFFER_SIZE / sizeof(u32))
>  #define CTB_G2H_BUFFER_SIZE	(SZ_128K)
> +#define CTB_G2H_BUFFER_DWORDS	(CTB_G2H_BUFFER_SIZE / sizeof(u32))
>  #define G2H_ROOM_BUFFER_SIZE	(CTB_G2H_BUFFER_SIZE / 2)
> +#define G2H_ROOM_BUFFER_DWORDS	(CTB_G2H_BUFFER_DWORDS / 2)
>  
>  /**
>   * xe_guc_ct_queue_proc_time_jiffies - Return maximum time to process a full
> @@ -403,7 +406,7 @@ int xe_guc_ct_init_post_hwconfig(struct xe_guc_ct *ct)
>  static void guc_ct_ctb_h2g_init(struct xe_device *xe, struct guc_ctb *h2g,
>  				struct iosys_map *map)
>  {
> -	h2g->info.size = CTB_H2G_BUFFER_SIZE / sizeof(u32);
> +	h2g->info.size = CTB_H2G_BUFFER_DWORDS;
>  	h2g->info.resv_space = 0;
>  	h2g->info.tail = 0;
>  	h2g->info.head = 0;
> @@ -421,8 +424,8 @@ static void guc_ct_ctb_h2g_init(struct xe_device *xe, struct guc_ctb *h2g,
>  static void guc_ct_ctb_g2h_init(struct xe_device *xe, struct guc_ctb *g2h,
>  				struct iosys_map *map)
>  {
> -	g2h->info.size = CTB_G2H_BUFFER_SIZE / sizeof(u32);
> -	g2h->info.resv_space = G2H_ROOM_BUFFER_SIZE / sizeof(u32);
> +	g2h->info.size = CTB_G2H_BUFFER_DWORDS;
> +	g2h->info.resv_space = G2H_ROOM_BUFFER_DWORDS;
>  	g2h->info.head = 0;
>  	g2h->info.tail = 0;
>  	g2h->info.space = CIRC_SPACE(g2h->info.tail, g2h->info.head,
> @@ -725,6 +728,12 @@ void xe_guc_ct_stop(struct xe_guc_ct *ct)
>   */
>  void xe_guc_ct_runtime_suspend(struct xe_guc_ct *ct)
>  {
> +	struct guc_ctb *g2h = &ct->ctbs.g2h;
> +	u32 credits = CIRC_SPACE(0, 0, CTB_G2H_BUFFER_DWORDS) - G2H_ROOM_BUFFER_DWORDS;
> +
> +	/* We should be back to guc_ct_ctb_g2h_init() values */
> +	xe_gt_assert(ct_to_gt(ct), g2h->info.space == credits);
> +
>  	/*
>  	 * Since we're already in runtime suspend path, we shouldn't have pending
>  	 * messages. But if there happen to be any, we'd probably want them to be
> -- 
> 2.43.0
> 

      parent reply	other threads:[~2025-12-02 16:01 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-12-02  8:33 [PATCH v1] drm/xe/guc_ct: Assert on credits mismatch during runtime suspend Raag Jadav
2025-12-02 10:33 ` ✓ CI.KUnit: success for " Patchwork
2025-12-02 11:34 ` ✓ Xe.CI.BAT: " Patchwork
2025-12-02 12:44 ` ✗ Xe.CI.Full: failure " Patchwork
2025-12-05  3:26   ` Raag Jadav
2025-12-02 16:01 ` Matthew Brost [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=aS8NZz36I8vOpWHn@lstrano-desk.jf.intel.com \
    --to=matthew.brost@intel.com \
    --cc=daniele.ceraolospurio@intel.com \
    --cc=intel-xe@lists.freedesktop.org \
    --cc=michal.wajdeczko@intel.com \
    --cc=raag.jadav@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