Wait for the whole boot prewarm and show its progress on a boot screen.

The game used to start after at most 3 s of prewarm, so after a translator, GPU or driver change it met the rest of
its shaders in play and stalled on each one. It now waits for all of it: a warm start is ready in about half a second
and shows nothing, while a cold one shows a "Preparing shaders" screen in the menu's colours and the game's fonts, with
the shader and then the pipeline count and a bar. The shader and pipeline prewarm report their numbers every 100 ms.
The present path takes a frame without a game image for the screen. The RomFS opens on the first host-side read, so
the screen's fonts load before the guest's first file access. NX2NAVI_PREWARM_WAIT_MS now defaults to 0 (no cap). With
no progress for 60 s the game starts anyway and the prewarm finishes in the background.

Measured with an emptied cache: 2,988 shaders translated and 2,399 pipelines built in 9.4 s behind the screen, then
0 ms shader and pipeline waits in Kokiri Forest. Warm: the game starts 0.47 s after launch.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
chrg6
2026-11-07 07:43:39 -05:00
parent 29fe2b4396
commit 375ecf27e2
12 changed files with 236 additions and 23 deletions

View File

@@ -1747,6 +1747,16 @@ typedef struct {
volatile int32_t missing, other_cached;
} Warm;
// The running prewarm's numbers for gfx_shader_prewarm_progress (boot.c's wait screen), copied from its Warm while it
// waits for the workers: the Warm lives on that function's stack.
static int64_t pw_done, pw_total, pw_xlated;
void gfx_shader_prewarm_progress(int64_t *done, int64_t *total, int64_t *translated) {
*done = __atomic_load_n(&pw_done, __ATOMIC_RELAXED);
*total = __atomic_load_n(&pw_total, __ATOMIC_RELAXED);
*translated = __atomic_load_n(&pw_xlated, __ATOMIC_RELAXED);
}
static void warm_worker(void *arg) {
Warm *w = arg;
for (;;) {
@@ -1822,6 +1832,9 @@ void gfx_shader_prewarm(const uint8_t (*iids)[32], size_t n, const GfxShader **o
w.n_other = other_inputs(sorted, sorted ? n : 0, &w.other);
free(sorted);
size_t total = w.n_named + w.n_other;
__atomic_store_n(&pw_done, 0, __ATOMIC_RELAXED);
__atomic_store_n(&pw_xlated, 0, __ATOMIC_RELAXED);
__atomic_store_n(&pw_total, (int64_t)total, __ATOMIC_RELAXED);
threads = threads < 1 ? 1 : threads > 64 ? 64 : threads;
if ((size_t)threads > total) threads = (int)(total ? total : 1);
svc_set_limit(procs);
@@ -1830,10 +1843,18 @@ void gfx_shader_prewarm(const uint8_t (*iids)[32], size_t n, const GfxShader **o
for (int i = 0; i < threads; i++)
if ((t[k] = plat_thread_new(warm_worker, &w, "shader prewarm"))) k++;
if (!k) warm_worker(&w);
double logged = (double)plat_now_ms();
for (int m = 0; m < k; m++)
while (!plat_thread_join_ns(t[m], 2000000000u))
while (!plat_thread_join_ns(t[m], 100000000u)) { // every 100 ms the numbers, every 2 s a log line
__atomic_store_n(&pw_done, (int64_t)w.done, __ATOMIC_RELAXED);
__atomic_store_n(&pw_xlated, (int64_t)((uint64_t)n_xlated - xl0), __ATOMIC_RELAXED);
if ((double)plat_now_ms() - logged < 2000) continue;
logged = (double)plat_now_ms();
rt_log("gfx_shader: prewarm: %lld of %zu shaders, %llu translated, %.0f ms", (long long)w.done, total,
(unsigned long long)((uint64_t)n_xlated - xl0), (double)plat_now_ms() - t0);
}
__atomic_store_n(&pw_done, (int64_t)total, __ATOMIC_RELAXED);
__atomic_store_n(&pw_xlated, (int64_t)((uint64_t)n_xlated - xl0), __ATOMIC_RELAXED);
svc_set_limit(svc_runtime_limit());
rt_log("gfx_shader: prewarm: %zu shaders named by pipeline keys (%ld not available), %zu other stored inputs (%ld "
"cached); %llu translated, %llu failed, with up to %d translators and %d threads, in %.0f ms", w.n_named,