Re: Parallel CREATE INDEX vs DSM starvation

2020-01-30 Thread Thomas Munro
On Thu, Jan 30, 2020 at 8:47 PM Peter Geoghegan wrote: > On Wed, Jan 29, 2020 at 11:38 PM Thomas Munro wrote: > > CreateParallelContext() can return a context with seg == NULL. That > > causes CREATE INDEX to segfault. Instead, it should fall back to > > non-parallel build. See attached. > > I

Re: Parallel CREATE INDEX vs DSM starvation

2020-01-29 Thread Peter Geoghegan
On Wed, Jan 29, 2020 at 11:38 PM Thomas Munro wrote: > CreateParallelContext() can return a context with seg == NULL. That > causes CREATE INDEX to segfault. Instead, it should fall back to > non-parallel build. See attached. I guess we can't call _bt_end_parallel() here. So your patch LGTM.

Parallel CREATE INDEX vs DSM starvation

2020-01-29 Thread Thomas Munro
Hello, CreateParallelContext() can return a context with seg == NULL. That causes CREATE INDEX to segfault. Instead, it should fall back to non-parallel build. See attached. This probably explains a segfault reported over on pgsql-general[1]. [1] https://www.postgresql.org/message-id/CA%2BhU