Hi Coly
>Hi Junhui,
>
>> btree_node_prefetch(b, k);
>> + /*
>> + * initiallize c->gc_stats.nodes
>> + * for incremental GC
>> + */
>> + b->c->gc_stats.nodes++;
>
>I don't get the point why increase gc_stats.nodes here. Could you please
>explain a bit more ?
We get gc_stats.nodes here, because we need to know the total btree nodes
to caculate how much btree nodes we could do incremental GC in the first
time. If we don't get it, it would be zero in the first GC time, and we
can't caculate how much btree nodes in incremental GC correctly in the
first GC time.
Though the count is not accurate (since after we get gc_stats.nodes, journal
replay, I/O writes may happend, so the real nodes count may be changed), but
we only use it to estimate how much btree nodes we could GC each time, so it
doesn't matter.
Thanks.
Tang Junhui
--
To unsubscribe from this list: send the line "unsubscribe linux-bcache" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
