| From: | Thomas Munro <thomas(dot)munro(at)gmail(dot)com> |
|---|---|
| To: | pgsql-hackers <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | More DSM slot exhaustion bugs |
| Date: | 2020-02-04 10:44:48 |
| Message-ID: | CA+hUKGKHa9ad=s-y-BOt0G+KsdwgHVTMrAzv7xMruzMgZwfSow@mail.gmail.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Hello,
Here are two more bugs I found by running the regression tests
repeatedly with dsm_create() hacked to fail at random.
1. If find_or_make_matching_shared_tupledesc() fails, we leave behind
a null pointer in RecordCacheHash, so that a later lookup segfaults.
2. If we do a rescan, then ExecHashJoinReInitializeDSM() needs to
return early if there is no DSM segment, otherwise a TOC lookup raises
a bogus error.
Here are some draft patches.
| Attachment | Content-Type | Size |
|---|---|---|
| 0001-Fix-shared-typmod-bug-on-DSM-slot-exhaustion.patch | application/octet-stream | 1.8 KB |
| 0002-Fix-parallel-hash-join-rescan-on-DSM-exhaustion.patch | application/octet-stream | 1.2 KB |
| 0003-Inject-random-faults-into-dsm_create.patch | application/octet-stream | 856 bytes |
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Rémi Lapeyre | 2020-02-04 13:25:03 | [PATCH v1] Allow COPY "text" format to output a header |
| Previous Message | Adam Lee | 2020-02-04 10:42:29 | Re: Memory-Bounded Hash Aggregation |