libc/dbm: Differentiate between uninitialized and end-of-db cursors
ClosedPublic
Actions

Authored by bnovkov on Fri, Jun 19, 3:41 PM.

Details

Reviewers

markj
ae

Commits

rGf879eb3febb6: libc/dbm: Differentiate between uninitialized and end-of-db cursors

Summary

Commit 3a686b851f8f fixed a dbm_nextkey edge case when using
the function after reaching the end of the database, but it inadvertently
broke the following R_NEXT behaviour:
"If the cursor is not yet set, this is the same as the R_FIRST flag."

Fix this by adding a new cursor constant that allows us to differentiate
between an unset cursor and a cursor that overflowed.

Reported by: ae
Fixes: 3a686b851f8f
Sponsored by: Klara, Inc.

Diff Detail

Repository

rG FreeBSD src repository

Lint

Lint Skipped

Unit

Tests Skipped

Build Status

Buildable 74013
Build 70896: arc lint + arc unit

Event Timeline

bnovkov created this revision.Fri, Jun 19, 3:41 PM

Herald added a subscriber: imp. · View Herald TranscriptFri, Jun 19, 3:41 PM

bnovkov requested review of this revision.Fri, Jun 19, 3:41 PM

Harbormaster completed remote builds in B74009: Diff 180071.Fri, Jun 19, 3:41 PM

bnovkov mentioned this in D51635: dbm_nextkey: Always return an error if we've reached the end of the database.Fri, Jun 19, 3:41 PM

markj added inline comments.Fri, Jun 19, 4:50 PM

lib/libc/db/hash/hash.c
722	The /* R_NEXT */ is kind of misleading now.
722	Are you sure that cbucket will never overflow? e.g., on line 771 or in hash_bigkey.c.

bnovkov added inline comments.Fri, Jun 19, 5:23 PM

lib/libc/db/hash/hash.c
722	Hm, it could overflow but the check on line 740 explicitly casts it back to `uint32_t` for the `MAX_BUCKET` comparison. I guess it wouldn't hurt if we turn the `==` into a `<=` though.