From: Matthew Brost <matthew.brost@intel.com>
To: Jonathan Cavitt <jonathan.cavitt@intel.com>
Cc: <intel-xe@lists.freedesktop.org>, <saurabhg.gupta@intel.com>,
<alex.zuo@intel.com>
Subject: Re: [PATCH] drm/xe/xe_guc_ct: Prevent compiler read/write optimization breaks
Date: Tue, 9 Dec 2025 16:32:39 -0800 [thread overview]
Message-ID: <aTi/p4KgMsWQLxW8@lstrano-desk.jf.intel.com> (raw)
In-Reply-To: <20251209212535.159493-2-jonathan.cavitt@intel.com>
On Tue, Dec 09, 2025 at 09:25:36PM +0000, Jonathan Cavitt wrote:
> Use READ_ONCE and WRITE_ONCE when operating on ct->state and the
> g2h_fence->done values to prevent the compiler from ignoring these
> necessary operations.
>
Maybe a fixes tag to guard against future compiler changes breaking
older kernels?
> Suggested-by: Matthew Brost <matthew.brost@intel.com>
> Signed-off-by: Jonathan Cavitt <jonathan.cavitt@intel.com>
> ---
> drivers/gpu/drm/xe/xe_guc_ct.c | 6 +++---
> drivers/gpu/drm/xe/xe_guc_ct.h | 4 ++--
> 2 files changed, 5 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/gpu/drm/xe/xe_guc_ct.c b/drivers/gpu/drm/xe/xe_guc_ct.c
> index 648f0f523abb..9129c50f5370 100644
> --- a/drivers/gpu/drm/xe/xe_guc_ct.c
> +++ b/drivers/gpu/drm/xe/xe_guc_ct.c
> @@ -206,7 +206,7 @@ static void g2h_fence_cancel(struct g2h_fence *g2h_fence)
> {
> g2h_fence->cancel = true;
> g2h_fence->fail = true;
> - g2h_fence->done = true;
> + WRITE_ONCE(g2h_fence->done, true);
Peer Thomas's feedback can we get pairs with comments around READ_ONCE /
WRITE_ONCE usages? Grep in Xe for READ_ONCE / WRITE_ONCE and you should
see comments that are an example of this.
Matt
> }
>
> static bool g2h_fence_needs_alloc(struct g2h_fence *g2h_fence)
> @@ -527,7 +527,7 @@ static void guc_ct_change_state(struct xe_guc_ct *ct,
> if (ct->g2h_outstanding)
> xe_pm_runtime_put(ct_to_xe(ct));
> ct->g2h_outstanding = 0;
> - ct->state = state;
> + WRITE_ONCE(ct->state, state);
>
> xe_gt_dbg(gt, "GuC CT communication channel %s\n",
> state == XE_GUC_CT_STATE_STOPPED ? "stopped" :
> @@ -1496,7 +1496,7 @@ static int parse_g2h_response(struct xe_guc_ct *ct, u32 *msg, u32 len)
>
> g2h_release_space(ct, GUC_CTB_HXG_MSG_MAX_LEN);
>
> - g2h_fence->done = true;
> + WRITE_ONCE(g2h_fence->done, true);
> smp_mb();
>
> wake_up_all(&ct->g2h_fence_wq);
> diff --git a/drivers/gpu/drm/xe/xe_guc_ct.h b/drivers/gpu/drm/xe/xe_guc_ct.h
> index 5599939f8fe1..cb1335f1d66f 100644
> --- a/drivers/gpu/drm/xe/xe_guc_ct.h
> +++ b/drivers/gpu/drm/xe/xe_guc_ct.h
> @@ -30,12 +30,12 @@ void xe_guc_ct_print(struct xe_guc_ct *ct, struct drm_printer *p, bool want_ctb)
>
> static inline bool xe_guc_ct_initialized(struct xe_guc_ct *ct)
> {
> - return ct->state != XE_GUC_CT_STATE_NOT_INITIALIZED;
> + return READ_ONCE(ct->state) != XE_GUC_CT_STATE_NOT_INITIALIZED;
> }
>
> static inline bool xe_guc_ct_enabled(struct xe_guc_ct *ct)
> {
> - return ct->state == XE_GUC_CT_STATE_ENABLED;
> + return READ_ONCE(ct->state) == XE_GUC_CT_STATE_ENABLED;
> }
>
> static inline void xe_guc_ct_irq_handler(struct xe_guc_ct *ct)
> --
> 2.43.0
>
next prev parent reply other threads:[~2025-12-10 0:32 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-12-09 21:25 [PATCH] drm/xe/xe_guc_ct: Prevent compiler read/write optimization breaks Jonathan Cavitt
2025-12-10 0:32 ` Matthew Brost [this message]
2025-12-10 13:55 ` ✓ CI.KUnit: success for " Patchwork
2025-12-10 15:04 ` ✓ Xe.CI.BAT: " Patchwork
2025-12-10 22:08 ` ✗ Xe.CI.Full: failure " Patchwork
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=aTi/p4KgMsWQLxW8@lstrano-desk.jf.intel.com \
--to=matthew.brost@intel.com \
--cc=alex.zuo@intel.com \
--cc=intel-xe@lists.freedesktop.org \
--cc=jonathan.cavitt@intel.com \
--cc=saurabhg.gupta@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