linux-kernel - Re: [PATCH-resent-to-correct-ml 2/8] drm/xe/gt: Unify xe_hw_fence_irq

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <qvwe46oojvk2njtotvsvvs6a4nbluic566mn7hykqc6uu2pdll@rrkvzms6jti3>
Date: Tue, 4 Feb 2025 09:20:24 -0600
From: Lucas De Marchi <lucas.demarchi@...el.com>
To: Maarten Lankhorst <dev@...khorst.se>
CC: <intel-xe@...ts.freedesktop.org>, <dri-devel@...ts.freedesktop.org>,
	<linux-kernel@...r.kernel.org>, Ingo Molnar <mingo@...nel.org>, David Lechner
	<dlechner@...libre.com>, Peter Zijlstra <peterz@...radead.org>, Will Deacon
	<will@...nel.org>, Waiman Long <longman@...hat.com>, Boqun Feng
	<boqun.feng@...il.com>
Subject: Re: [PATCH-resent-to-correct-ml 2/8] drm/xe/gt: Unify
 xe_hw_fence_irq_finish() calls.

On Tue, Feb 04, 2025 at 02:22:31PM +0100, Maarten Lankhorst wrote:
>Those calls should be from xe_gt_init, not the diverse amount of places
>they are called.
>
>Signed-off-by: Maarten Lankhorst <dev@...khorst.se>
>---
> drivers/gpu/drm/xe/xe_gt.c | 31 ++++++++++++++-----------------
> 1 file changed, 14 insertions(+), 17 deletions(-)
>
>diff --git a/drivers/gpu/drm/xe/xe_gt.c b/drivers/gpu/drm/xe/xe_gt.c
>index 01a4a852b8f43..943bab94119fa 100644
>--- a/drivers/gpu/drm/xe/xe_gt.c
>+++ b/drivers/gpu/drm/xe/xe_gt.c
>@@ -408,13 +408,11 @@ static void dump_pat_on_error(struct xe_gt *gt)
> static int gt_fw_domain_init(struct xe_gt *gt)
> {
> 	unsigned int fw_ref;
>-	int err, i;
>+	int err;
>
> 	fw_ref = xe_force_wake_get(gt_to_fw(gt), XE_FW_GT);
>-	if (!fw_ref) {
>-		err = -ETIMEDOUT;
>-		goto err_hw_fence_irq;
>-	}
>+	if (!fw_ref)
>+		return -ETIMEDOUT;
>
> 	if (!xe_gt_is_media_type(gt)) {
> 		err = xe_ggtt_init(gt_to_tile(gt)->mem.ggtt);
>@@ -455,9 +453,6 @@ static int gt_fw_domain_init(struct xe_gt *gt)
> err_force_wake:
> 	dump_pat_on_error(gt);
> 	xe_force_wake_put(gt_to_fw(gt), fw_ref);
>-err_hw_fence_irq:
>-	for (i = 0; i < XE_ENGINE_CLASS_MAX; ++i)
>-		xe_hw_fence_irq_finish(&gt->fence_irq[i]);
>
> 	return err;
> }
>@@ -465,7 +460,7 @@ static int gt_fw_domain_init(struct xe_gt *gt)
> static int all_fw_domain_init(struct xe_gt *gt)
> {
> 	unsigned int fw_ref;
>-	int err, i;
>+	int err;
>
> 	fw_ref = xe_force_wake_get(gt_to_fw(gt), XE_FORCEWAKE_ALL);
> 	if (!xe_force_wake_ref_has_domain(fw_ref, XE_FORCEWAKE_ALL)) {
>@@ -543,8 +538,6 @@ static int all_fw_domain_init(struct xe_gt *gt)
>
> err_force_wake:
> 	xe_force_wake_put(gt_to_fw(gt), fw_ref);
>-	for (i = 0; i < XE_ENGINE_CLASS_MAX; ++i)
>-		xe_hw_fence_irq_finish(&gt->fence_irq[i]);
>
> 	return err;
> }

the two parts above should be fine and is similar to what I've sent on
my probe cleanup:

https://lore.kernel.org/intel-xe/20250131223140.4144292-7-lucas.demarchi@intel.com/


>@@ -596,35 +589,39 @@ int xe_gt_init(struct xe_gt *gt)
>
> 	err = xe_gt_pagefault_init(gt);
> 	if (err)
>-		return err;
>+		goto err;
>
> 	xe_mocs_init_early(gt);
>
> 	err = xe_gt_sysfs_init(gt);
> 	if (err)
>-		return err;
>+		goto err;
>
> 	err = gt_fw_domain_init(gt);
> 	if (err)
>-		return err;
>+		goto err;
>
> 	err = xe_gt_idle_init(&gt->gtidle);
> 	if (err)
>-		return err;
>+		goto err;
>
> 	err = xe_gt_freq_init(gt);
> 	if (err)
>-		return err;
>+		goto err;
>
> 	xe_force_wake_init_engines(gt, gt_to_fw(gt));
>
> 	err = all_fw_domain_init(gt);
> 	if (err)
>-		return err;
>+		goto err;
>
> 	xe_gt_record_user_engines(gt);
>
> 	return 0;
>+err:
>+	for (i = 0; i < XE_ENGINE_CLASS_MAX; ++i)
>+		xe_hw_fence_irq_finish(&gt->fence_irq[i]);
>+	return err;

this however is moving to the opposite direction from

https://lore.kernel.org/intel-xe/20250131223140.4144292-6-lucas.demarchi@intel.com/

Lucas De Marchi

> }
>
> /**
>-- 
>2.47.1
>