On Mon, 2020-09-07 at 08:49 +0000, Johannes Thumshirn wrote: > On 07/09/2020 05:06, Damien Le Moal wrote: > > May be you meant something like "leave a zone not active after a truncate when > > the zone file is open for writing" ? > > No I meant, we shouldn't decrement the 's_open_zones' count on truncate to 0 > or full, when a zone is still opened for write. Because if we do, another thread > could open the last available open zone and the application won't be able to write > to a previously opened zone, if that makes sense. I understood that, but the "active writers" is hard to decode/understand. So may be just replace that with "as the file may still be open for writing, e.g. the user called ftruncate(). If the zone file is not open and a process does a truncate(), then no close operation is needed." -- Damien Le Moal Western Digital Research
