MacKwnoX/NX2navi
1
0
Fork 0
mirror of https://github.com/tangerineturtle1/NX2navi.git synced 2026-11-15 19:19:58 -04:00
Code Issues Packages Projects Releases Wiki Activity

Name the queue lock's long holders under NX2NAVI_LOCK_TIME, and time each swapchain rebuild's hold of it.

Browse Source 
With lock_time = 1 the 0.10.0 session showed flip_locked waiting up to 48.61 ms for the queue lock, and the replay's
queue_submit up to 8.05 ms. Nothing named who held it: only vk_texture.c's lock records its holders. A swapchain
rebuild holds the lock across vkDeviceWaitIdle and the create, and that hold was not timed either.

vk_queue_lock, vk_queue_lock_raw and vk_queue_trylock_raw are no longer inlined. Each takes the lock with
plat_rmutex_lock_at and its own return address, so the wait report keeps naming the caller (before, the inlined call
named a site inside flip_locked). At the outermost take, with lock_time on, the holder's site and the time are noted.
vk_queue_unlock passes a hold of 5 ms or more to perf.c perf_qlock_hold. The frame-time summary then prints "queue lock
held 5 ms or more" beside the texture lock's holds, by site, as the waits are. create_swapchain logs how long a
successful rebuild took: the wait for the lock, the hold, and the part spent waiting for the device to idle. It logs
the first 16 rebuilds and each doubling of their count.

Diagnostics only: the counters are behind NX2NAVI_LOCK_TIME, which is off by default, plus one capped line per
rebuild. Checks that passed: pass_cc_check on every runtime file (host.h changed), no new -Wall -Wextra warnings in
gfx_present.c and perf.c, stutter_hitch_check, pgo_build_test, sync_hw_check.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
chrg6
2026-11-13 17:21:28 -05:00
parent 68bc197393
commit 0ffac80222
3 changed files with 57 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

10
runtime/perf.c
View File

@@ -516,6 +516,7 @@ static uint64_t summary_t0; // start of the current period (0: not started)
#define LT_N 512
static struct { void *site; uint64_t ns, n, max, ns0, n0; } lt[LT_N];
static struct { void *site; uint64_t ns, n, max, ns0, n0; } lh[LT_N]; // texture lock holds over 1 ms (perf_lock_hold)
static struct { void *site; uint64_t ns, n, max, ns0, n0; } lq[LT_N]; // queue lock holds of 5 ms or more (perf_qlock_hold)
static uint64_t lt_lost_ns, lt_rec0[2];
static volatile int64_t fr_hold_ns; // the texture lock's holds over 1 ms since the last present (ft_frame)
void (*perf_rec_waits)(uint64_t *full_ns, uint64_t *drain_ns);
@@ -541,6 +542,7 @@ void perf_lock_hold(void *site, uint64_t ns) {
lt_add((typeof(lt[0]) *)lh, site, ns);
plat_add64(&fr_hold_ns, (int64_t)ns);
}
void perf_qlock_hold(void *site, uint64_t ns) { lt_add((typeof(lt[0]) *)lq, site, ns); }
static int lt_on = -1;
int perf_lock_holds_on(void) {
@@ -582,7 +584,8 @@ static uint64_t lt_top(typeof(lt[0]) *tab, char *line, int *n, size_t cap, uint6
return all;
}
// The lock lines of a summary: the replay threads' waits with the recorder's, then the texture lock's long holds.
// The lock lines of a summary: the replay threads' waits with the recorder's, then the texture lock's and the queue
// lock's long holds.
static void lt_report(void) {
uint64_t rec[2] = {0, 0}, waits, holds;
if (perf_rec_waits) perf_rec_waits(&rec[0], &rec[1]);
@@ -598,6 +601,11 @@ static void lt_report(void) {
snprintf(line, sizeof line, "frametime: texture lock held over 1 ms: %.0f ms (%llu holds):%s", all / 1e6,
(unsigned long long)holds, top);
rt_log("%s", line);
n = 0, top[0] = 0;
all = lt_top((typeof(lt[0]) *)lq, top, &n, sizeof top, &holds);
snprintf(line, sizeof line, "frametime: queue lock held 5 ms or more: %.0f ms (%llu holds):%s", all / 1e6,
(unsigned long long)holds, top);
rt_log("%s", line);
lt_rec0[0] = rec[0], lt_rec0[1] = rec[1], lt_lost_ns = 0;
}