[fix](be) Fix cross-thread CLOCK_THREAD_CPUTIME_ID read in scanner CPU timer

### What problem does this PR solve?

Problem Summary: Scanner::_cpu_watch (ThreadCpuStopWatch using CLOCK_THREAD_CPUTIME_ID)
was started via resume() on a scanner worker thread but read via pause() on the pipeline
task thread. Since CLOCK_THREAD_CPUTIME_ID is a per-thread CPU clock, reading it on a
different thread produces garbage/negative values, triggering the DCHECK:

  Check failed: _value.load() > -1L (-39943795 vs. -1) delta: -252570258

In the non-EOS path of _scanner_scan(), update_scanner_profile() (which calls pause())
was only called for the EOS path. The non-EOS path left _cpu_watch running and later
ScannerScheduler::submit() called pause() from the pipeline task thread.

Fix:
1. Always call update_scanner_profile() before push_back_scan_task() in _scanner_scan(),
   ensuring pause() runs on the scanner worker thread for both EOS and non-EOS paths.
2. Reinitialize _cpu_watch after reading in _update_scan_cpu_timer() so that any
   subsequent cross-thread pause() call in submit() safely reads 0.

### Release note

None

### Check List (For Author)

- Test: Manual test - verified the logic by code analysis
- Behavior changed: No
- Does this need documentation: No

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>

Author: jacktengg

be/src/exec/scan/scanner.cpp

@@ -279,6 +279,11 @@ void Scanner::_collect_profile_before_close() {
 
 void Scanner::_update_scan_cpu_timer() {
     int64_t cpu_time = _cpu_watch.elapsed_time();
+    // Reinitialize _cpu_watch after reading. _cpu_watch uses CLOCK_THREAD_CPUTIME_ID
+    // (per-thread CPU clock), so it must only be read on the thread that started it.
+    // Reinitializing here prevents stale cross-thread reads if pause() is called again
+    // from a different thread (e.g., in ScannerScheduler::submit() on the pipeline thread).
+    _cpu_watch = ThreadCpuStopWatch();
     _scan_cpu_timer += cpu_time;
     if (_state && _state->get_query_ctx()) {
         _state->get_query_ctx()->resource_ctx()->cpu_context()->update_cpu_cost_ms(cpu_time);

be/src/exec/scan/scanner_scheduler.cpp

@@ -301,10 +301,12 @@ void ScannerScheduler::_scanner_scan(std::shared_ptr<ScannerContext> ctx,
         eos = true;
     }
 
+    // Always update scanner profile (including pause()) on the scanner worker thread
+    // before leaving. _cpu_watch uses CLOCK_THREAD_CPUTIME_ID which is per-thread,
+    // so elapsed_time() must be read on the same thread that called resume()/start().
+    update_scanner_profile();
+
     if (eos) {
-        // If eos, scanner will call _collect_profile_before_close to update profile,
-        // so we need update_scanner_profile here
-        update_scanner_profile();
         scanner->mark_to_need_to_close();
         scan_task->set_state(ScanTask::State::EOS);
     } else {