Speed up the replay and fix what the perf pass found, and add a frame capture key for chasing rendering faults.

Performance: draws reuse the bindings that did not change and record only the dynamic states that differ from what
the command buffer holds; small uploads come from a staging arena; finished submissions are held briefly and handed to
the queue in batches; the recorder thread can keep a CPU core of its own; ASTC block-linear textures are deswizzled on
the GPU; dynamic resolution steps the upscaler mode to hold a frame-time target. Each has a setting in nx2navi.ini
(bind_reuse, dyn_delta, stg_arena, submit_batch, recorder_core, recorder_prio, astc_gpu_deswizzle, dyn_res) and the
staging cache now frees what it holds at shutdown.

Capture: NX2NAVI_DUMP=key saves the frame after each press of Insert, with mesh draws listed by task, mesh and fragment
shader, depth test and face culling, and mesh_prims.txt giving each mesh draw's primitive, clipping and fragment counts.
All of it is off unless NX2NAVI_DUMP is set.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
chrg6
2026-11-09 01:35:47 -05:00
parent 86de3320a0
commit 1f7e9c975b
38 changed files with 3105 additions and 1840 deletions

View File

@@ -390,6 +390,29 @@ int plat_thread_join_ns(PlatThread *t, uint64_t ns) {
// The thread keeps using t->fn and t->arg, so the struct stays allocated (a few bytes per detached thread).
void plat_thread_detach(PlatThread *t) { pthread_detach(t->t); }
// The calling thread gets the last processor of its affinity mask to itself: every other thread of the process now
// running loses that processor (threads they start later inherit their masks), and the caller keeps only it. Returns
// 1, or 0 when the mask has fewer than three processors or the caller's mask cannot be set.
int plat_cpu_reserve(void) {
cpu_set_t all, mine, rest;
if (sched_getaffinity(0, sizeof all, &all) || CPU_COUNT(&all) < 3) return 0;
int last = -1;
for (int i = CPU_SETSIZE - 1; i >= 0 && last < 0; i--)
if (CPU_ISSET(i, &all)) last = i;
rest = all;
CPU_CLR(last, &rest);
CPU_ZERO(&mine);
CPU_SET(last, &mine);
pid_t me = (pid_t)syscall(SYS_gettid);
DIR *d = opendir("/proc/self/task");
for (struct dirent *e; d && (e = readdir(d));) {
pid_t tid = (pid_t)atoi(e->d_name);
if (tid > 0 && tid != me) sched_setaffinity(tid, sizeof rest, &rest);
}
if (d) closedir(d);
return sched_setaffinity(0, sizeof mine, &mine) == 0;
}
uint64_t plat_thread_affinity(PlatThread *t, uint64_t mask) {
pthread_t th = t ? t->t : pthread_self();
cpu_set_t old, set;