Redis sdslen() for addReply(c,shared.crlf) is taking a 6.55% of the total CPU cycles for large array replies

Using the lrange benchmarks of redis benchmark ( I was investigating it due to #10981 ) we can see that we're wasting a large ammount of cpu cycles on sdslen() computing the sizes we already know:

                           - 77.55% lrangeCommand                                                                                                                                                          ▒
                              - 77.31% addListRangeReply                                                                                                                                                   ▒
                                 + 35.46% addReplyBulkCBuffer                                                                                                                                              ▒
                                 + 21.34% listTypeNext (inlined)                                                                                                                                           ▒
                                 - 11.32% addReply (inlined)                                                                                                                                               ▒
                                      6.55% sdslen (inlined)                                                                                                                                               ▒
                                      0.41% _addReplyToBufferOrList (inlined)                                                                                                                              ▒
                                 + 6.88% _addReplyToBufferOrList (inlined)                                                                                                                                 ▒
                                 + 0.60% addReplyLongLongWithPrefix                                                                                                                                        ▒
                                   0.50% prepareClientToWrite                                                                                                                                              ▒
                                   0.15% listTypeReleaseIterator (inlined)