Am 13.06.24 um 08:47 schrieb Sebastian Andrzej Siewior:
PREEMPT_RT has a different locking implementation for ww_mutex. The
base mutex of struct ww_mutex is declared as struct WW_MUTEX_BASE. The
latter is defined as `mutex' for non-PREEMPT_RT builds and `rt_mutex'
for PREEMPT_RT builds.
Using mutex_lock() directly on the base mutex in
ttm_bo_reserve_deadlock() leads to compile error on PREEMPT_RT.
The locking-selftest has its own defines to deal with this and it is
probably best to defines the needed one within the test program since
their usefulness is limited outside of well known selftests.
Provide ww_mutex_base_lock() which points to the correct function for
PREEMPT_RT and non-PREEMPT_RT builds.
In general good that somebody is looking into this, but I can't even
judge why ww_mutex would use rt_mutex in the first place.
So I don't feel well reviewing this.
Regards,
Christian.
Fixes: 995279d280d1e ("drm/ttm/tests: Add tests for ttm_bo functions")
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@xxxxxxxxxxxxx>
---
Repost of https://lore.kernel.org/r/20240404102534.QTa80QPY@xxxxxxxxxxxxx
drivers/gpu/drm/ttm/tests/ttm_bo_test.c | 8 +++++++-
1 file changed, 7 insertions(+), 1 deletion(-)
--- a/drivers/gpu/drm/ttm/tests/ttm_bo_test.c
+++ b/drivers/gpu/drm/ttm/tests/ttm_bo_test.c
@@ -18,6 +18,12 @@
#define BO_SIZE SZ_8K
+#ifdef CONFIG_PREEMPT_RT
+#define ww_mutex_base_lock(b) rt_mutex_lock(b)
+#else
+#define ww_mutex_base_lock(b) mutex_lock(b)
+#endif
+
struct ttm_bo_test_case {
const char *description;
bool interruptible;
@@ -142,7 +148,7 @@ static void ttm_bo_reserve_deadlock(stru
bo2 = ttm_bo_kunit_init(test, test->priv, BO_SIZE);
ww_acquire_init(&ctx1, &reservation_ww_class);
- mutex_lock(&bo2->base.resv->lock.base);
+ ww_mutex_base_lock(&bo2->base.resv->lock.base);
/* The deadlock will be caught by WW mutex, don't warn about it */
lock_release(&bo2->base.resv->lock.base.dep_map, 1);