postgresql – Impact on the performance of updating target columns with the same values ​​IN CONFLICT

This is a question about the internal operation of Postgres (v10) and performance.

Given a table, github_repos, with a unique index of several columns in org_id Y github_id columns, is there any difference in performance (or other issues to be taken into account) between the two volume increase operations below? The difference is that in the first consultation, the org_id Y github_id the columns are included in the UPDATE, while in the second consultation they are not included. Since the UPDATE is executed in CONFLICT, the updated values ​​of org_id Y github_id they will be the same as the previous values, but those columns are included in the UPDATE because the underlying library that I am using is designed that way. I wonder if it is safe to use it as it is or if it should be explicitly excluding those columns in the UPDATE.

Query # 1:

INSERT IN "github_repos" ("org_id", "github_id", "name")
VALUES (1,1, & # 39; foo & # 39;)
IN CONFLICT (org_id, github_id)
MAKE UPDATE CONFIGURE "org_id" = EXCLUDE. "Org_id", "github_id" = EXCLUDES. "Github_id", "name" = EXCLUDED. "First name"
RETURN "ID"

Query # 2:

INSERT IN "github_repos" ("org_id", "github_id", "name")
VALUES (1,1, & # 39; foo & # 39;)
IN CONFLICT (org_id, github_id)
UPDATE CONFIGURE "name" = EXCLUDED. "First name"
RETURN "ID"

github_repos table:

                        Column | Type | Collation | Voidable
------------------- + ------------------- + ---------- - + ---------- +
id | bigint | | not null |
org_id | bigint | | not null |
github_id | bigint | | not null |
name | Character that varies | | not null |

Indexes:
"github_repos_pkey" PRIMARY KEY, btree (id)
"unique_repos" UNIQUE, btree (org_id, github_id)