Trace logical addresses when operating on partition

[ksztyber]

Trace the logical address instead of the physical one when operating on a partition of a device.
For instance, the logs below show a write to the first block within the /dev/nvme0n1p1 partition, while the traces (collected from the partition device, /dev/nvme0n1p1, not /dev/nvme0n1) indicate a write being sent to address 2048 (start of the /dev/nvme0n1p1 partition). It seems more logical to record the traces offset within the bounds of the partition if user specifies the partition as the device to be traced.

[root@iotrace iotrace]# fdisk -l /dev/nvme0n1

Units = sectors of 1 * 4096 = 4096 bytes
Sector size (logical/physical): 4096 bytes / 4096 bytes
I/O size (minimum/optimal): 4096 bytes / 4096 bytes
Disk label type: dos
Disk identifier: 0x3c1bcb0d

        Device Boot      Start         End      Blocks   Id  System
/dev/nvme0n1p1             256     7078143    28311552   83  Linux

[root@iotrace iotrace]# iotrace -S -d /dev/nvme0n1p1 &
[1] 58924
[root@iotrace iotrace]# dd if=/dev/zero of=/dev/nvme0n1p1 bs=4k count=1 seek=0 oflag=dsync
1+0 records in
1+0 records out
4096 bytes (4.1 kB) copied, 0.000327407 s, 12.5 MB/s

[root@iotrace iotrace]# kill 58924
...
{
 "tracePath": "kernel/2019-08-20_05:41:32",
 "state": "COMPLETE",
 "sourceNode": {
  "node": [
   {
    "id": "kernel"
   }
  ]
 },
 "traceStartDateTime": "2019-08-20 05:41",
 "traceDuration": "7",
 "traceSize": 1,
 "tracedEvents": "61",
 "droppedEvents": "0",
 "queueCount": "36"
}

[root@iotrace iotrace]# iotrace --parse-trace --path kernel/2019-08-20_05:41:32 | grep Write
{"header":{"sid":"60473671","timestamp":"525077100448357"},"io":{"lba":"2048","len":8,"ioClass":1,"deviceId":"271581184","operation":"Write","flush":false,"fua":false}}