From 34bb7b813ab398106f700b0a6b218509bb0b904c Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Thomas=20Hellstr=C3=B6m?= <thomas.hellstrom@linux.intel.com>
Date: Tue, 3 Sep 2024 11:42:32 +0200
Subject: [PATCH] drm/xe: Use xe_pm_runtime_get in xe_bo_move() if
reclaim-safe.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
xe_bo_move() might be called in the TTM swapout path from validation
by another TTM device. If so, we are not likely to have a RPM
reference. So iff xe_pm_runtime_get() is safe to call from reclaim,
use it instead of xe_pm_runtime_get_noresume().
Strictly this is currently needed only if handle_system_ccs is true,
but use xe_pm_runtime_get() if possible anyway to increase test
coverage.
At the same time warn if handle_system_ccs is true and we can't
call xe_pm_runtime_get() from reclaim context. This will likely trip
if someone tries to enable SRIOV on LNL, without fixing Xe SRIOV
runtime resume / suspend.
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Matthew Brost <matthew.brost@intel.com>
Cc: Matthew Auld <matthew.auld@intel.com>
Signed-off-by: Thomas Hellström <thomas.hellstrom@linux.intel.com>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240903094232.166342-1-thomas.hellstrom@linux.intel.com
---
drivers/gpu/drm/xe/xe_bo.c | 11 ++++++++++-
drivers/gpu/drm/xe/xe_pm.c | 9 ++++++++-
drivers/gpu/drm/xe/xe_pm.h | 1 +
3 files changed, 19 insertions(+), 2 deletions(-)
diff --git a/drivers/gpu/drm/xe/xe_bo.c b/drivers/gpu/drm/xe/xe_bo.c
index 1c18ba9bd0996..0c64d3b3155e8 100644
--- a/drivers/gpu/drm/xe/xe_bo.c
+++ b/drivers/gpu/drm/xe/xe_bo.c
@@ -758,7 +758,16 @@ static int xe_bo_move(struct ttm_buffer_object *ttm_bo, bool evict,
xe_assert(xe, migrate);
trace_xe_bo_move(bo, new_mem->mem_type, old_mem_type, move_lacks_source);
- xe_pm_runtime_get_noresume(xe);
+ if (xe_rpm_reclaim_safe(xe)) {
+ /*
+ * We might be called through swapout in the validation path of
+ * another TTM device, so unconditionally acquire rpm here.
+ */
+ xe_pm_runtime_get(xe);
+ } else {
+ drm_WARN_ON(&xe->drm, handle_system_ccs);
+ xe_pm_runtime_get_noresume(xe);
+ }
if (xe_bo_is_pinned(bo) && !xe_bo_is_user(bo)) {
/*
diff --git a/drivers/gpu/drm/xe/xe_pm.c b/drivers/gpu/drm/xe/xe_pm.c
index 2600c936527e1..e518557e0eec6 100644
--- a/drivers/gpu/drm/xe/xe_pm.c
+++ b/drivers/gpu/drm/xe/xe_pm.c
@@ -79,7 +79,14 @@ static struct lockdep_map xe_pm_runtime_nod3cold_map = {
};
#endif
-static bool __maybe_unused xe_rpm_reclaim_safe(const struct xe_device *xe)
+/**
+ * xe_rpm_reclaim_safe() - Whether runtime resume can be done from reclaim context
+ * @xe: The xe device.
+ *
+ * Return: true if it is safe to runtime resume from reclaim context.
+ * false otherwise.
+ */
+bool xe_rpm_reclaim_safe(const struct xe_device *xe)
{
return !xe->d3cold.capable && !xe->info.has_sriov;
}
diff --git a/drivers/gpu/drm/xe/xe_pm.h b/drivers/gpu/drm/xe/xe_pm.h
index 9aef673b1c8ad..998d1ed645560 100644
--- a/drivers/gpu/drm/xe/xe_pm.h
+++ b/drivers/gpu/drm/xe/xe_pm.h
@@ -31,6 +31,7 @@ bool xe_pm_runtime_resume_and_get(struct xe_device *xe);
void xe_pm_assert_unbounded_bridge(struct xe_device *xe);
int xe_pm_set_vram_threshold(struct xe_device *xe, u32 threshold);
void xe_pm_d3cold_allowed_toggle(struct xe_device *xe);
+bool xe_rpm_reclaim_safe(const struct xe_device *xe);
struct task_struct *xe_pm_read_callback_task(struct xe_device *xe);
int xe_pm_module_init(void);
--
GitLab