On Sun, 2023-02-26 at 22:29 +0800, jack...@gmail.com wrote: > > here are the source codes from src/include/access/htup_details.h. > /* > * information stored in t_infomask: > */ > #define HEAP_HASNULL 0x0001 /* has null attribute(s) */ > #define HEAP_HASVARWIDTH 0x0002 /* has variable-width > attribute(s) */ > #define HEAP_HASEXTERNAL 0x0004 /* has external stored > attribute(s) */ > #define HEAP_HASOID_OLD 0x0008 /* has an object-id > field */ > #define HEAP_XMAX_KEYSHR_LOCK 0x0010 /* xmax is a key-shared locker */ > #define HEAP_COMBOCID 0x0020 /* t_cid is a combo CID */ > #define HEAP_XMAX_EXCL_LOCK 0x0040 /* xmax is exclusive locker */ > #define HEAP_XMAX_LOCK_ONLY 0x0080 /* xmax, if valid, is only a > locker */ > > And I can't understand these attrs: > 1. external stored attribute(s), what is this? can you give a create > statement to show me?
Those are columns stored out of line. See https://www.postgresql.org/docs/current/storage-toast.html > 2. xmax is a key-shared locker/exclusive locker/only a locker, so how you use > this? can you give me a scenario? https://www.cybertec-postgresql.com/en/whats-in-an-xmax/ > let me try to explain it: > if there is a txn is trying to read this heaptuple, the > HEAP_XMAX_KEYSHR_LOCK bit will be set to 1. No, reading a row won't lock it. It's SELECT ... FOR KEY SHARE that does that. > if there is a txn is trying to delete/update this heaptuple, the > HEAP_XMAX_EXCL_LOCK bit will be set to 1. Correct. > but for HEAP_XMAX_LOCK_ONLY, I can't understand. See the comment a few source lines below: /* * A tuple is only locked (i.e. not updated by its Xmax) if the * HEAP_XMAX_LOCK_ONLY bit is set; or, for pg_upgrade's sake, if the Xmax is * not a multi and the EXCL_LOCK bit is set. Essentially, it is set after SELECT ... FOR UPDATE. > And another thought is that these three bit can have only one to be set 1 at > most. If a tuple is only locked FOR KEY SHARE, bit not updated or deleted, both are set. > 3. t_cid is a combo CID? what's a CID? give me an example please. The command ID of the command in the transaction. Yours, Laurenz Albe
