summaryrefslogtreecommitdiff
path: root/drivers/gpu/drm/i915/gem/selftests/i915_gem_context.c
diff options
context:
space:
mode:
Diffstat (limited to 'drivers/gpu/drm/i915/gem/selftests/i915_gem_context.c')
-rw-r--r--drivers/gpu/drm/i915/gem/selftests/i915_gem_context.c227
1 files changed, 124 insertions, 103 deletions
diff --git a/drivers/gpu/drm/i915/gem/selftests/i915_gem_context.c b/drivers/gpu/drm/i915/gem/selftests/i915_gem_context.c
index 3f41fe5ec9d4..1330c0b431a7 100644
--- a/drivers/gpu/drm/i915/gem/selftests/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/gem/selftests/i915_gem_context.c
@@ -5,9 +5,12 @@
*/
#include <linux/prime_numbers.h>
+#include <linux/string_helpers.h>
+#include "gem/i915_gem_internal.h"
#include "gem/i915_gem_pm.h"
#include "gt/intel_engine_pm.h"
+#include "gt/intel_engine_regs.h"
#include "gt/intel_gt.h"
#include "gt/intel_gt_requests.h"
#include "gt/intel_reset.h"
@@ -63,7 +66,7 @@ static int live_nop_switch(void *arg)
ctx[n] = live_context(i915, file);
if (IS_ERR(ctx[n])) {
err = PTR_ERR(ctx[n]);
- goto out_file;
+ goto out_ctx;
}
}
@@ -79,7 +82,7 @@ static int live_nop_switch(void *arg)
this = igt_request_alloc(ctx[n], engine);
if (IS_ERR(this)) {
err = PTR_ERR(this);
- goto out_file;
+ goto out_ctx;
}
if (rq) {
i915_request_await_dma_fence(this, &rq->fence);
@@ -90,10 +93,10 @@ static int live_nop_switch(void *arg)
}
if (i915_request_wait(rq, 0, 10 * HZ) < 0) {
pr_err("Failed to populated %d contexts\n", nctx);
- intel_gt_set_wedged(to_gt(i915));
+ intel_gt_set_wedged(engine->gt);
i915_request_put(rq);
err = -EIO;
- goto out_file;
+ goto out_ctx;
}
i915_request_put(rq);
@@ -104,7 +107,7 @@ static int live_nop_switch(void *arg)
err = igt_live_test_begin(&t, i915, __func__, engine->name);
if (err)
- goto out_file;
+ goto out_ctx;
end_time = jiffies + i915_selftest.timeout_jiffies;
for_each_prime_number_from(prime, 2, 8192) {
@@ -117,7 +120,7 @@ static int live_nop_switch(void *arg)
this = igt_request_alloc(ctx[n % nctx], engine);
if (IS_ERR(this)) {
err = PTR_ERR(this);
- goto out_file;
+ goto out_ctx;
}
if (rq) { /* Force submission order */
@@ -146,7 +149,7 @@ static int live_nop_switch(void *arg)
if (i915_request_wait(rq, 0, HZ / 5) < 0) {
pr_err("Switching between %ld contexts timed out\n",
prime);
- intel_gt_set_wedged(to_gt(i915));
+ intel_gt_set_wedged(engine->gt);
i915_request_put(rq);
break;
}
@@ -162,7 +165,7 @@ static int live_nop_switch(void *arg)
err = igt_live_test_end(&t);
if (err)
- goto out_file;
+ goto out_ctx;
pr_info("Switch latencies on %s: 1 = %lluns, %lu = %lluns\n",
engine->name,
@@ -170,103 +173,116 @@ static int live_nop_switch(void *arg)
prime - 1, div64_u64(ktime_to_ns(times[1]), prime - 1));
}
+out_ctx:
+ kfree(ctx);
out_file:
fput(file);
return err;
}
struct parallel_switch {
- struct task_struct *tsk;
+ struct kthread_worker *worker;
+ struct kthread_work work;
struct intel_context *ce[2];
+ int result;
};
-static int __live_parallel_switch1(void *data)
+static void __live_parallel_switch1(struct kthread_work *work)
{
- struct parallel_switch *arg = data;
+ struct parallel_switch *arg =
+ container_of(work, typeof(*arg), work);
IGT_TIMEOUT(end_time);
unsigned long count;
count = 0;
+ arg->result = 0;
do {
struct i915_request *rq = NULL;
- int err, n;
+ int n;
- err = 0;
- for (n = 0; !err && n < ARRAY_SIZE(arg->ce); n++) {
+ for (n = 0; !arg->result && n < ARRAY_SIZE(arg->ce); n++) {
struct i915_request *prev = rq;
rq = i915_request_create(arg->ce[n]);
if (IS_ERR(rq)) {
i915_request_put(prev);
- return PTR_ERR(rq);
+ arg->result = PTR_ERR(rq);
+ break;
}
i915_request_get(rq);
if (prev) {
- err = i915_request_await_dma_fence(rq, &prev->fence);
+ arg->result =
+ i915_request_await_dma_fence(rq,
+ &prev->fence);
i915_request_put(prev);
}
i915_request_add(rq);
}
- if (i915_request_wait(rq, 0, HZ / 5) < 0)
- err = -ETIME;
+
+ if (IS_ERR_OR_NULL(rq))
+ break;
+
+ if (i915_request_wait(rq, 0, HZ) < 0)
+ arg->result = -ETIME;
+
i915_request_put(rq);
- if (err)
- return err;
count++;
- } while (!__igt_timeout(end_time, NULL));
+ } while (!arg->result && !__igt_timeout(end_time, NULL));
- pr_info("%s: %lu switches (sync)\n", arg->ce[0]->engine->name, count);
- return 0;
+ pr_info("%s: %lu switches (sync) <%d>\n",
+ arg->ce[0]->engine->name, count, arg->result);
}
-static int __live_parallel_switchN(void *data)
+static void __live_parallel_switchN(struct kthread_work *work)
{
- struct parallel_switch *arg = data;
+ struct parallel_switch *arg =
+ container_of(work, typeof(*arg), work);
struct i915_request *rq = NULL;
IGT_TIMEOUT(end_time);
unsigned long count;
int n;
count = 0;
+ arg->result = 0;
do {
- for (n = 0; n < ARRAY_SIZE(arg->ce); n++) {
+ for (n = 0; !arg->result && n < ARRAY_SIZE(arg->ce); n++) {
struct i915_request *prev = rq;
- int err = 0;
rq = i915_request_create(arg->ce[n]);
if (IS_ERR(rq)) {
i915_request_put(prev);
- return PTR_ERR(rq);
+ arg->result = PTR_ERR(rq);
+ break;
}
i915_request_get(rq);
if (prev) {
- err = i915_request_await_dma_fence(rq, &prev->fence);
+ arg->result =
+ i915_request_await_dma_fence(rq,
+ &prev->fence);
i915_request_put(prev);
}
i915_request_add(rq);
- if (err) {
- i915_request_put(rq);
- return err;
- }
}
count++;
- } while (!__igt_timeout(end_time, NULL));
- i915_request_put(rq);
+ } while (!arg->result && !__igt_timeout(end_time, NULL));
- pr_info("%s: %lu switches (many)\n", arg->ce[0]->engine->name, count);
- return 0;
+ if (!IS_ERR_OR_NULL(rq))
+ i915_request_put(rq);
+
+ pr_info("%s: %lu switches (many) <%d>\n",
+ arg->ce[0]->engine->name, count, arg->result);
}
static int live_parallel_switch(void *arg)
{
struct drm_i915_private *i915 = arg;
- static int (* const func[])(void *arg) = {
+ static void (* const func[])(struct kthread_work *) = {
__live_parallel_switch1,
__live_parallel_switchN,
NULL,
@@ -274,7 +290,7 @@ static int live_parallel_switch(void *arg)
struct parallel_switch *data = NULL;
struct i915_gem_engines *engines;
struct i915_gem_engines_iter it;
- int (* const *fn)(void *arg);
+ void (* const *fn)(struct kthread_work *);
struct i915_gem_context *ctx;
struct intel_context *ce;
struct file *file;
@@ -332,8 +348,10 @@ static int live_parallel_switch(void *arg)
continue;
ce = intel_context_create(data[m].ce[0]->engine);
- if (IS_ERR(ce))
+ if (IS_ERR(ce)) {
+ err = PTR_ERR(ce);
goto out;
+ }
err = intel_context_pin(ce);
if (err) {
@@ -345,9 +363,24 @@ static int live_parallel_switch(void *arg)
}
}
+ for (n = 0; n < count; n++) {
+ struct kthread_worker *worker;
+
+ if (!data[n].ce[0])
+ continue;
+
+ worker = kthread_run_worker(0, "igt/parallel:%s",
+ data[n].ce[0]->engine->name);
+ if (IS_ERR(worker)) {
+ err = PTR_ERR(worker);
+ goto out;
+ }
+
+ data[n].worker = worker;
+ }
+
for (fn = func; !err && *fn; fn++) {
struct igt_live_test t;
- int n;
err = igt_live_test_begin(&t, i915, __func__, "");
if (err)
@@ -357,34 +390,23 @@ static int live_parallel_switch(void *arg)
if (!data[n].ce[0])
continue;
- data[n].tsk = kthread_run(*fn, &data[n],
- "igt/parallel:%s",
- data[n].ce[0]->engine->name);
- if (IS_ERR(data[n].tsk)) {
- err = PTR_ERR(data[n].tsk);
- break;
- }
- get_task_struct(data[n].tsk);
+ data[n].result = 0;
+ kthread_init_work(&data[n].work, *fn);
+ kthread_queue_work(data[n].worker, &data[n].work);
}
- yield(); /* start all threads before we kthread_stop() */
-
for (n = 0; n < count; n++) {
- int status;
-
- if (IS_ERR_OR_NULL(data[n].tsk))
- continue;
-
- status = kthread_stop(data[n].tsk);
- if (status && !err)
- err = status;
-
- put_task_struct(data[n].tsk);
- data[n].tsk = NULL;
+ if (data[n].ce[0]) {
+ kthread_flush_work(&data[n].work);
+ if (data[n].result && !err)
+ err = data[n].result;
+ }
}
- if (igt_live_test_end(&t))
- err = -EIO;
+ if (igt_live_test_end(&t)) {
+ err = err ?: -EIO;
+ break;
+ }
}
out:
@@ -396,6 +418,9 @@ out:
intel_context_unpin(data[n].ce[m]);
intel_context_put(data[n].ce[m]);
}
+
+ if (data[n].worker)
+ kthread_destroy_worker(data[n].worker);
}
kfree(data);
out_file:
@@ -452,7 +477,8 @@ static int gpu_fill(struct intel_context *ce,
static int cpu_fill(struct drm_i915_gem_object *obj, u32 value)
{
const bool has_llc = HAS_LLC(to_i915(obj->base.dev));
- unsigned int n, m, need_flush;
+ unsigned int need_flush;
+ unsigned long n, m;
int err;
i915_gem_object_lock(obj, NULL);
@@ -463,12 +489,12 @@ static int cpu_fill(struct drm_i915_gem_object *obj, u32 value)
for (n = 0; n < real_page_count(obj); n++) {
u32 *map;
- map = kmap_atomic(i915_gem_object_get_page(obj, n));
+ map = kmap_local_page(i915_gem_object_get_page(obj, n));
for (m = 0; m < DW_PER_PAGE; m++)
map[m] = value;
if (!has_llc)
drm_clflush_virt_range(map, PAGE_SIZE);
- kunmap_atomic(map);
+ kunmap_local(map);
}
i915_gem_object_finish_access(obj);
@@ -482,7 +508,8 @@ out:
static noinline int cpu_check(struct drm_i915_gem_object *obj,
unsigned int idx, unsigned int max)
{
- unsigned int n, m, needs_flush;
+ unsigned int needs_flush;
+ unsigned long n;
int err;
i915_gem_object_lock(obj, NULL);
@@ -491,15 +518,15 @@ static noinline int cpu_check(struct drm_i915_gem_object *obj,
goto out_unlock;
for (n = 0; n < real_page_count(obj); n++) {
- u32 *map;
+ u32 *map, m;
- map = kmap_atomic(i915_gem_object_get_page(obj, n));
+ map = kmap_local_page(i915_gem_object_get_page(obj, n));
if (needs_flush & CLFLUSH_BEFORE)
drm_clflush_virt_range(map, PAGE_SIZE);
for (m = 0; m < max; m++) {
if (map[m] != m) {
- pr_err("%pS: Invalid value at object %d page %d/%ld, offset %d/%d: found %x expected %x\n",
+ pr_err("%pS: Invalid value at object %d page %ld/%ld, offset %d/%d: found %x expected %x\n",
__builtin_return_address(0), idx,
n, real_page_count(obj), m, max,
map[m], m);
@@ -510,7 +537,7 @@ static noinline int cpu_check(struct drm_i915_gem_object *obj,
for (; m < DW_PER_PAGE; m++) {
if (map[m] != STACK_MAGIC) {
- pr_err("%pS: Invalid value at object %d page %d, offset %d: found %x expected %x (uninitialised)\n",
+ pr_err("%pS: Invalid value at object %d page %ld, offset %d: found %x expected %x (uninitialised)\n",
__builtin_return_address(0), idx, n, m,
map[m], STACK_MAGIC);
err = -EINVAL;
@@ -519,7 +546,7 @@ static noinline int cpu_check(struct drm_i915_gem_object *obj,
}
out_unmap:
- kunmap_atomic(map);
+ kunmap_local(map);
if (err)
break;
}
@@ -698,7 +725,7 @@ static int igt_ctx_exec(void *arg)
pr_err("Failed to fill dword %lu [%lu/%lu] with gpu (%s) [full-ppgtt? %s], err=%d\n",
ndwords, dw, max_dwords(obj),
engine->name,
- yesno(i915_gem_context_has_full_ppgtt(ctx)),
+ str_yes_no(i915_gem_context_has_full_ppgtt(ctx)),
err);
intel_context_put(ce);
kernel_context_close(ctx);
@@ -832,7 +859,7 @@ static int igt_shared_ctx_exec(void *arg)
pr_err("Failed to fill dword %lu [%lu/%lu] with gpu (%s) [full-ppgtt? %s], err=%d\n",
ndwords, dw, max_dwords(obj),
engine->name,
- yesno(i915_gem_context_has_full_ppgtt(ctx)),
+ str_yes_no(i915_gem_context_has_full_ppgtt(ctx)),
err);
intel_context_put(ce);
kernel_context_close(ctx);
@@ -883,7 +910,9 @@ out_file:
return err;
}
-static int rpcs_query_batch(struct drm_i915_gem_object *rpcs, struct i915_vma *vma)
+static int rpcs_query_batch(struct drm_i915_gem_object *rpcs,
+ struct i915_vma *vma,
+ struct intel_engine_cs *engine)
{
u32 *cmd;
@@ -894,9 +923,9 @@ static int rpcs_query_batch(struct drm_i915_gem_object *rpcs, struct i915_vma *v
return PTR_ERR(cmd);
*cmd++ = MI_STORE_REGISTER_MEM_GEN8;
- *cmd++ = i915_mmio_reg_offset(GEN8_R_PWR_CLK_STATE);
- *cmd++ = lower_32_bits(vma->node.start);
- *cmd++ = upper_32_bits(vma->node.start);
+ *cmd++ = i915_mmio_reg_offset(GEN8_R_PWR_CLK_STATE(engine->mmio_base));
+ *cmd++ = lower_32_bits(i915_vma_offset(vma));
+ *cmd++ = upper_32_bits(i915_vma_offset(vma));
*cmd = MI_BATCH_BUFFER_END;
__i915_gem_object_flush_map(rpcs, 0, 64);
@@ -933,13 +962,14 @@ emit_rpcs_query(struct drm_i915_gem_object *obj,
if (IS_ERR(rpcs))
return PTR_ERR(rpcs);
+ i915_gem_ww_ctx_init(&ww, false);
+
batch = i915_vma_instance(rpcs, ce->vm, NULL);
if (IS_ERR(batch)) {
err = PTR_ERR(batch);
goto err_put;
}
- i915_gem_ww_ctx_init(&ww, false);
retry:
err = i915_gem_object_lock(obj, &ww);
if (!err)
@@ -955,7 +985,7 @@ retry:
if (err)
goto err_vma;
- err = rpcs_query_batch(rpcs, vma);
+ err = rpcs_query_batch(rpcs, vma, ce->engine);
if (err)
goto err_batch;
@@ -965,15 +995,11 @@ retry:
goto err_batch;
}
- err = i915_request_await_object(rq, batch->obj, false);
- if (err == 0)
- err = i915_vma_move_to_active(batch, rq, 0);
+ err = i915_vma_move_to_active(batch, rq, 0);
if (err)
goto skip_request;
- err = i915_request_await_object(rq, vma->obj, true);
- if (err == 0)
- err = i915_vma_move_to_active(vma, rq, EXEC_OBJECT_WRITE);
+ err = i915_vma_move_to_active(vma, rq, EXEC_OBJECT_WRITE);
if (err)
goto skip_request;
@@ -984,7 +1010,8 @@ retry:
}
err = rq->engine->emit_bb_start(rq,
- batch->node.start, batch->node.size,
+ i915_vma_offset(batch),
+ i915_vma_size(batch),
0);
if (err)
goto skip_request;
@@ -1374,7 +1401,7 @@ static int igt_ctx_readonly(void *arg)
goto out_file;
}
- vm = ctx->vm ?: &i915->ggtt.alias->vm;
+ vm = ctx->vm ?: &to_gt(i915)->ggtt->alias->vm;
if (!vm || !vm->has_read_only) {
err = 0;
goto out_file;
@@ -1411,7 +1438,7 @@ static int igt_ctx_readonly(void *arg)
pr_err("Failed to fill dword %lu [%lu/%lu] with gpu (%s) [full-ppgtt? %s], err=%d\n",
ndwords, dw, max_dwords(obj),
ce->engine->name,
- yesno(i915_gem_context_has_full_ppgtt(ctx)),
+ str_yes_no(i915_gem_context_has_full_ppgtt(ctx)),
err);
i915_gem_context_unlock_engines(ctx);
goto out_file;
@@ -1533,11 +1560,7 @@ static int write_to_scratch(struct i915_gem_context *ctx,
goto err_unpin;
}
- i915_vma_lock(vma);
- err = i915_request_await_object(rq, vma->obj, false);
- if (err == 0)
- err = i915_vma_move_to_active(vma, rq, 0);
- i915_vma_unlock(vma);
+ err = igt_vma_move_to_active_unlocked(vma, rq, 0);
if (err)
goto skip_request;
@@ -1547,7 +1570,8 @@ static int write_to_scratch(struct i915_gem_context *ctx,
goto skip_request;
}
- err = engine->emit_bb_start(rq, vma->node.start, vma->node.size, 0);
+ err = engine->emit_bb_start(rq, i915_vma_offset(vma),
+ i915_vma_size(vma), 0);
if (err)
goto skip_request;
@@ -1652,7 +1676,7 @@ static int read_from_scratch(struct i915_gem_context *ctx,
*cmd++ = offset;
*cmd++ = MI_STORE_REGISTER_MEM | MI_USE_GGTT;
*cmd++ = reg;
- *cmd++ = vma->node.start + result;
+ *cmd++ = i915_vma_offset(vma) + result;
*cmd = MI_BATCH_BUFFER_END;
i915_gem_object_flush_map(obj);
@@ -1669,11 +1693,7 @@ static int read_from_scratch(struct i915_gem_context *ctx,
goto err_unpin;
}
- i915_vma_lock(vma);
- err = i915_request_await_object(rq, vma->obj, true);
- if (err == 0)
- err = i915_vma_move_to_active(vma, rq, EXEC_OBJECT_WRITE);
- i915_vma_unlock(vma);
+ err = igt_vma_move_to_active_unlocked(vma, rq, EXEC_OBJECT_WRITE);
if (err)
goto skip_request;
@@ -1683,7 +1703,8 @@ static int read_from_scratch(struct i915_gem_context *ctx,
goto skip_request;
}
- err = engine->emit_bb_start(rq, vma->node.start, vma->node.size, flags);
+ err = engine->emit_bb_start(rq, i915_vma_offset(vma),
+ i915_vma_size(vma), flags);
if (err)
goto skip_request;
generated by cgit (git 2.34.1) at 2025-12-27 01:03:19 +0000