@@ -577,14 +577,14 @@ void submit_sliding_window1d(const PaddedSpan<const T, SizeT> &a,
577
577
// auto *const out_end = out.end();
578
578
579
579
auto y_start = glid;
580
- auto y_stop = std::min (y_start + WorkPI*results.size_x (), out.size ());
580
+ auto y_stop =
581
+ std::min (y_start + WorkPI * results.size_x (), out.size ());
581
582
int32_t i = 0 ;
582
- for (uint32_t y = y_start; y < y_stop; y+=results.size_x ())
583
- {
583
+ for (uint32_t y = y_start; y < y_stop; y += results.size_x ()) {
584
584
out_ptr[y] = results[i++];
585
585
}
586
- // due to excessive optimizations this code results in memory corruption
587
- // results.store(&out_ptr[glid],
586
+ // due to excessive optimizations this code results in memory
587
+ // corruption results.store(&out_ptr[glid],
588
588
// [out_end](auto &&ptr) { return ptr < out_end; });
589
589
});
590
590
}
@@ -647,14 +647,14 @@ void submit_sliding_window1d_small_kernel(const PaddedSpan<const T, SizeT> &a,
647
647
// auto *const out_end = out.end();
648
648
649
649
auto y_start = glid;
650
- auto y_stop = std::min (y_start + WorkPI*results.size_x (), out.size ());
650
+ auto y_stop =
651
+ std::min (y_start + WorkPI * results.size_x (), out.size ());
651
652
int32_t i = 0 ;
652
- for (uint32_t y = y_start; y < y_stop; y+=results.size_x ())
653
- {
653
+ for (uint32_t y = y_start; y < y_stop; y += results.size_x ()) {
654
654
out_ptr[y] = results[i++];
655
655
}
656
- // due to excessive optimizations this code results in memory corruption
657
- // results.store(&out_ptr[glid],
656
+ // due to excessive optimizations this code results in memory
657
+ // corruption results.store(&out_ptr[glid],
658
658
// [out_end](auto &&ptr) { return ptr < out_end; });
659
659
});
660
660
}
0 commit comments