lib/{mergeset,storage}: avoid unaligned 64-bit atomic operation panic on 32-bit platforms

The panic has been introduced in 68f3a02589

While at it, add padding to shard structs in order to avoid false sharing on mordern CPUs

This should improve scalability on systems with many CPU cores

lib/mergeset/table.go

@@ -10,6 +10,7 @@ import (
 	"sync"
 	"sync/atomic"
 	"time"
+	"unsafe"
 
 	"github.com/VictoriaMetrics/VictoriaMetrics/lib/bytesutil"
 	"github.com/VictoriaMetrics/VictoriaMetrics/lib/cgroup"
@@ -158,10 +159,20 @@ func (riss *rawItemsShards) Len() int {
 	return n
 }
 
-type rawItemsShard struct {
+type rawItemsShardNopad struct {
+	// Put lastFlushTime to the top in order to avoid unaligned memory access on 32-bit architectures
+	lastFlushTime uint64
+
 	mu  sync.Mutex
 	ibs []*inmemoryBlock
-	lastFlushTime uint64
+}
+
+type rawItemsShard struct {
+	rawItemsShardNopad
+
+	// The padding prevents false sharing on widespread platforms with
+	// 128 mod (cache line size) = 0 .
+	_ [128 - unsafe.Sizeof(rawItemsShardNopad{})%128]byte
 }
 
 func (ris *rawItemsShard) Len() int {

lib/storage/partition.go

@@ -445,10 +445,20 @@ func (rrss *rawRowsShards) Len() int {
 	return n
 }
 
-type rawRowsShard struct {
+type rawRowsShardNopad struct {
+	// Put lastFlushTime to the top in order to avoid unaligned memory access on 32-bit architectures
+	lastFlushTime uint64
+
 	mu   sync.Mutex
 	rows []rawRow
-	lastFlushTime uint64
+}
+
+type rawRowsShard struct {
+	rawRowsShardNopad
+
+	// The padding prevents false sharing on widespread platforms with
+	// 128 mod (cache line size) = 0 .
+	_ [128 - unsafe.Sizeof(rawRowsShardNopad{})%128]byte
 }
 
 func (rrs *rawRowsShard) Len() int {