fix: restore safe erasure read buffer pooling

Wire the preallocated pooled shard slices into parallelReader instead
of discarding them and allocating a buffer for each disk.

Keep readerToBuf as a permutation while preferred readers are reordered.
The former assignments could duplicate a buffer slot after multiple
swaps, causing concurrent writes and a possible decode stall. Add pool
aliasing and mapping regression coverage.

Co-authored-by: ChatGPT <noreply@openai.com>
Co-authored-by: Claude <noreply@anthropic.com>

cmd/erasure-decode.go

@@ -69,7 +69,7 @@ func newParallelReader(readers []io.ReaderAt, e Erasure, offset, totalLength int
 		offset:        (offset / e.blockSize) * e.ShardSize(),
 		shardSize:     e.ShardSize(),
 		shardFileSize: e.ShardFileSize(totalLength),
-		buf:           make([][]byte, len(readers)),
+		buf:           bufs,
 		readerToBuf:   r2b,
 		stashBuffer:   b,
 	}
@@ -106,8 +106,7 @@ func (p *parallelReader) preferReaders(prefer []bool) {
 		// Move reader with index i to index next.
 		// Do this by swapping next and i
 		p.readers[next], p.readers[i] = p.readers[i], p.readers[next]
-		p.readerToBuf[next] = i
-		p.readerToBuf[i] = next
+		p.readerToBuf[next], p.readerToBuf[i] = p.readerToBuf[i], p.readerToBuf[next]
 		next++
 	}
 }