1. The log_slow_filter value accepted by the server is qc_miss rather than qc_hit. Which is inconsistent with the "QC_Hit:" value printed in the slow log, but we have to live with that as we cannot change it in stable releases. However, the documentation for log_slow_filter is now incorrect.
2. The bug (more precisely, item(2) from the report) applies to both 5.1 and 5.5. I'm going to target it to 5.1 series.
Couple of comments:
1. The log_slow_filter value accepted by the server is qc_miss rather than qc_hit. Which is inconsistent with the "QC_Hit:" value printed in the slow log, but we have to live with that as we cannot change it in stable releases. However, the documentation for log_slow_filter is now incorrect.
2. The bug (more precisely, item(2) from the report) applies to both 5.1 and 5.5. I'm going to target it to 5.1 series.