[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <CAAofZF5iwRcEg9813aZf+Bib36u3K9MSjCwt478rzxuKW97gcw@mail.gmail.com>
Date: Thu, 8 Jan 2026 17:53:56 +0100
From: Marco Crivellari <marco.crivellari@...e.com>
To: linux-kernel@...r.kernel.org, intel-xe@...ts.freedesktop.org,
dri-devel@...ts.freedesktop.org
Cc: Tejun Heo <tj@...nel.org>, Lai Jiangshan <jiangshanlai@...il.com>,
Frederic Weisbecker <frederic@...nel.org>, Sebastian Andrzej Siewior <bigeasy@...utronix.de>,
Michal Hocko <mhocko@...e.com>, Lucas De Marchi <lucas.demarchi@...el.com>,
Thomas Hellstrom <thomas.hellstrom@...ux.intel.com>, Rodrigo Vivi <rodrigo.vivi@...el.com>,
David Airlie <airlied@...il.com>, Simona Vetter <simona@...ll.ch>
Subject: Re: [PATCH v2 2/2] drm/xe: add WQ_PERCPU to alloc_workqueue users
Hi,
I realized something in the meantime, looking manually at the patch,
while preparing the v3.
There is a bug in the alloc_workqueue() line in xe_ggtt_init_early():
On Mon, Nov 3, 2025 at 6:06 PM Marco Crivellari
<marco.crivellari@...e.com> wrote:
> [...]
> diff --git a/drivers/gpu/drm/xe/xe_ggtt.c b/drivers/gpu/drm/xe/xe_ggtt.c
> index 5edc0cad47e2..566163ab96ae 100644
> --- a/drivers/gpu/drm/xe/xe_ggtt.c
> +++ b/drivers/gpu/drm/xe/xe_ggtt.c
> @@ -291,7 +291,7 @@ int xe_ggtt_init_early(struct xe_ggtt *ggtt)
> else
> ggtt->pt_ops = &xelp_pt_ops;
>
> - ggtt->wq = alloc_workqueue("xe-ggtt-wq", 0, WQ_MEM_RECLAIM);
> + ggtt->wq = alloc_workqueue("xe-ggtt-wq", WQ_PERCPU, WQ_MEM_RECLAIM);
> if (!ggtt->wq)
> return -ENOMEM;
> - ggtt->wq = alloc_workqueue("xe-ggtt-wq", 0, WQ_MEM_RECLAIM);
The workqueue flag is the 2nd parameter, not the 3rd. The 3rd is max_active.
I will send a patch in order to fix this before proceeding with this
entire series.
Thanks!
--
Marco Crivellari
L3 Support Engineer
Powered by blists - more mailing lists