Re: [libvirt] [PATCH 1/3] qemu: monitor: return block stats data as a hash to avoid disk mixup

On 09/25/2014 06:06 AM, Peter Krempa wrote: > The current block stats code matched up the disk name with the actual > stats by the order in the data returned from qemu. This unfortunately > isn't right as qemu may return the disks in any order. Fix this by > returning a hash of stats and index them by the disk alias. > --- > src/qemu/qemu_driver.c | 44 ++++++++--------- > src/qemu/qemu_monitor.c | 14 ++---- > src/qemu/qemu_monitor.h | 6 +-- > src/qemu/qemu_monitor_json.c | 112 ++++++++++++++++++++----------------------- > src/qemu/qemu_monitor_json.h | 4 +- > 5 files changed, 82 insertions(+), 98 deletions(-) > > @@ -1749,56 +1750,62 @@ int qemuMonitorJSONGetBlockStatsInfo(qemuMonitorPtr mon, > if (flush_total_times) > *flush_total_times = -1; > > - if (qemuMonitorJSONGetAllBlockStatsInfo(mon, dev_name, &stats, 1) != 1) > + if (qemuMonitorJSONGetAllBlockStatsInfo(mon, &blockstats) < 0) > goto cleanup; > > - *rd_req = stats.rd_req; > - *rd_bytes = stats.rd_bytes; > - *wr_req = stats.wr_req; > - *wr_bytes = stats.wr_bytes; > + if (!(stats = virHashLookup(blockstats, dev_name))) { Maybe slightly less efficient (we now malloc stats for all members, only to throw away everything but the one we want), but I can live with it. ACK; worth having in 1.2.9