Intel-XE Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Matthew Brost <matthew.brost@intel.com>
To: Brian Nguyen <brian3.nguyen@intel.com>
Cc: <intel-xe@lists.freedesktop.org>, <stuart.summers@intel.com>,
	<matthew.d.roper@intel.com>
Subject: Re: [PATCH v2] drm/xe: Implement print vfunc for page fault info logging
Date: Mon, 28 Sep 2026 13:53:47 -0700	[thread overview]
Message-ID: <arrT2/evj15NQc/X@gsse-cloud1.jf.intel.com> (raw)
In-Reply-To: <20260928183120.3788244-2-brian3.nguyen@intel.com>

On Mon, Sep 28, 2026 at 06:31:21PM +0000, Brian Nguyen wrote:
> Add a print vfunc to xe_pagefault_ops so different producer backends can
> log specific relevant debug info while keeping the consumer layer free
> of potential backend-specific formatting logic.
> 
> Consumer-specific error prints are handled by passing as input into
> the vfunc.
> 
> v2:
>  - Decode producer specific fields for print message. (Matthew)
>  - Combine consumer and producer prints to one print call. (Sashiko)
> 
> Suggested-by: Stuart Summers <stuart.summers@intel.com>
> Signed-off-by: Brian Nguyen <brian3.nguyen@intel.com>
> Cc: Matthew Brost <matthew.brost@intel.com>
> ---
>  drivers/gpu/drm/xe/xe_guc_pagefault.c   | 32 ++++++++++++++++++++++++
>  drivers/gpu/drm/xe/xe_pagefault.c       | 33 +++----------------------
>  drivers/gpu/drm/xe/xe_pagefault_types.h | 10 ++++++++
>  3 files changed, 45 insertions(+), 30 deletions(-)
> 
> diff --git a/drivers/gpu/drm/xe/xe_guc_pagefault.c b/drivers/gpu/drm/xe/xe_guc_pagefault.c
> index df237fd40551..e65e4962a26c 100644
> --- a/drivers/gpu/drm/xe/xe_guc_pagefault.c
> +++ b/drivers/gpu/drm/xe/xe_guc_pagefault.c
> @@ -4,9 +4,11 @@
>   */
>  
>  #include "abi/guc_actions_abi.h"
> +#include "xe_gt_printk.h"
>  #include "xe_guc.h"
>  #include "xe_guc_ct.h"
>  #include "xe_guc_pagefault.h"
> +#include "xe_hw_engine.h"
>  #include "xe_pagefault.h"
>  #include "xe_pagefault_types.h"
>  
> @@ -68,10 +70,40 @@ static void guc_ack_fault_end(void *private)
>  	xe_guc_ct_unlock(&guc->ct);
>  }
>  
> +static void xe_guc_pagefault_print(struct xe_pagefault *pf,
> +				   const char *err_str)
> +{
> +	const u32 *msg = pf->producer.msg;
> +	u32 engine_class = FIELD_GET(PFD_ENG_CLASS, msg[0]);
> +
> +	xe_gt_info(pf->gt, "\n\tASID: %lu\n"
> +		   "\tFaulted Address: 0x%08lx%08lx\n"
> +		   "\tFaultType: %lu\n"
> +		   "\tAccessType: %lu\n"
> +		   "\tFaultLevel: %lu\n"
> +		   "\tEngineClass: %u %s\n"
> +		   "\tEngineInstance: %lu\n"
> +		   "\tSRCID: 0x%02lx\n"
> +		   "\tError: %s\n",
> +		   FIELD_GET(PFD_ASID, msg[1]),
> +		   FIELD_GET(PFD_VIRTUAL_ADDR_HI, msg[3]),
> +		   FIELD_GET(PFD_VIRTUAL_ADDR_LO, msg[2]) <<
> +		   PFD_VIRTUAL_ADDR_LO_SHIFT,
> +		   FIELD_GET(PFD_FAULT_TYPE, msg[2]),
> +		   FIELD_GET(PFD_ACCESS_TYPE, msg[2]),
> +		   FIELD_GET(PFD_FAULT_LEVEL, msg[0]),
> +		   engine_class,
> +		   xe_hw_engine_class_to_str(engine_class),
> +		   FIELD_GET(PFD_ENG_INSTANCE, msg[0]),
> +		   FIELD_GET(PFD_SRC_ID, msg[0]),
> +		   err_str);
> +}
> +
>  static const struct xe_pagefault_ops guc_pagefault_ops = {
>  	.ack_fault_begin = guc_ack_fault_begin,
>  	.ack_fault = guc_ack_fault,
>  	.ack_fault_end = guc_ack_fault_end,
> +	.print = xe_guc_pagefault_print,
>  };
>  
>  /**
> diff --git a/drivers/gpu/drm/xe/xe_pagefault.c b/drivers/gpu/drm/xe/xe_pagefault.c
> index f219e9e3b73e..dbf9397938df 100644
> --- a/drivers/gpu/drm/xe/xe_pagefault.c
> +++ b/drivers/gpu/drm/xe/xe_pagefault.c
> @@ -13,7 +13,6 @@
>  #include "xe_gt_printk.h"
>  #include "xe_gt_types.h"
>  #include "xe_gt_stats.h"
> -#include "xe_hw_engine.h"
>  #include "xe_log.h"
>  #include "xe_migrate.h"
>  #include "xe_pagefault.h"
> @@ -610,36 +609,10 @@ static const char *xe_pagefault_error_to_str(enum xe_pagefault_error error)
>  
>  static void xe_pagefault_print(struct xe_pagefault *pf)
>  {
> -	u8 engine_class = FIELD_GET(XE_PAGEFAULT_ENGINE_CLASS_MASK,
> -				    pf->consumer.engine_class_instance);
> -	u64 addr = xe_pagefault_addr(pf);
> +	enum xe_pagefault_error err = xe_pagefault_get_error(pf);
>  
> -	xe_gt_info(pf->gt, "\n\tASID: %lu\n"
> -		   "\tFaulted Address: 0x%08x%08x\n"
> -		   "\tFaultType: %lu\n"
> -		   "\tAccessType: %lu\n"
> -		   "\tFaultLevel: %lu\n"
> -		   "\tEngineClass: %d %s\n"
> -		   "\tEngineInstance: %lu\n"
> -		   "\tSRCID: 0x%02lx\n"
> -		   "\tError: %s\n",
> -		   FIELD_GET(XE_PAGEFAULT_ASID_MASK,
> -			     pf->consumer.id),
> -		   upper_32_bits(addr),
> -		   lower_32_bits(addr),
> -		   FIELD_GET(XE_PAGEFAULT_TYPE_MASK,
> -			     pf->consumer.fault_type_level),
> -		   FIELD_GET(XE_PAGEFAULT_ACCESS_TYPE_MASK,
> -			     pf->consumer.access_type),
> -		   FIELD_GET(XE_PAGEFAULT_LEVEL_MASK,
> -			     pf->consumer.fault_type_level),
> -		   engine_class,
> -		   xe_hw_engine_class_to_str(engine_class),
> -		   FIELD_GET(XE_PAGEFAULT_ENGINE_INSTANCE_MASK,
> -			     pf->consumer.engine_class_instance),
> -		   FIELD_GET(XE_PAGEFAULT_SRCID_MASK,
> -			     pf->consumer.id),
> -		   xe_pagefault_error_to_str(xe_pagefault_get_error(pf)));
> +	xe_gt_assert(pf->gt, pf->producer.ops->print);

AI reviews will likely tell you this useless as we'd NULL ptr deref
immediate after this - I got some of those on my code for similar type
of asserts.

I'd say this is correct, so how about...

if (WARN_ON_ONCE(!pf->producer.ops->print))
	return;

pf->producer.ops->print(pf, xe_pagefault_error_to_str(err));

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

> +	pf->producer.ops->print(pf, xe_pagefault_error_to_str(err));
>  }
>  
>  static void xe_pagefault_save_to_vm(struct xe_device *xe, struct xe_pagefault *pf)
> diff --git a/drivers/gpu/drm/xe/xe_pagefault_types.h b/drivers/gpu/drm/xe/xe_pagefault_types.h
> index 8ae6b9848bdc..ee467117afe9 100644
> --- a/drivers/gpu/drm/xe/xe_pagefault_types.h
> +++ b/drivers/gpu/drm/xe/xe_pagefault_types.h
> @@ -139,6 +139,16 @@ struct xe_pagefault_ops {
>  	 * Page fault producer ends acknowledgment from the consumer.
>  	 */
>  	void (*ack_fault_end)(void *private);
> +	/**
> +	 * @print: Print pagefault info
> +	 * @pf: Page fault
> +	 * @err_str: Consumer servicing error string
> +	 *
> +	 * Print pagefault identification fields decoded from producer,
> +	 * along with consumer-specific @err_str prints.
> +	 * Must be implemented by all producers.
> +	 */
> +	void (*print)(struct xe_pagefault *pf, const char *err_str);
>  };
>  
>  /**
> -- 
> 2.43.0
> 

  parent reply	other threads:[~2026-09-28 20:53 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-09-28 18:31 [PATCH v2] drm/xe: Implement print vfunc for page fault info logging Brian Nguyen
2026-09-28 19:38 ` ✓ CI.KUnit: success for drm/xe: Implement print vfunc for page fault info logging (rev2) Patchwork
2026-09-28 20:53 ` ✓ Xe.CI.BAT: " Patchwork
2026-09-28 20:53 ` Matthew Brost [this message]
2026-09-28 22:02   ` [PATCH v2] drm/xe: Implement print vfunc for page fault info logging Nguyen, Brian

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=arrT2/evj15NQc/X@gsse-cloud1.jf.intel.com \
    --to=matthew.brost@intel.com \
    --cc=brian3.nguyen@intel.com \
    --cc=intel-xe@lists.freedesktop.org \
    --cc=matthew.d.roper@intel.com \
    --cc=stuart.summers@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