colexec: cfetcher errors on setup when indexed column contains unhandled type#42999
Merged
craig[bot] merged 1 commit intocockroachdb:masterfrom Dec 5, 2019
Merged
Conversation
Member
|
This change is |
yuzefovich
approved these changes
Dec 5, 2019
Member
yuzefovich
left a comment
There was a problem hiding this comment.
cc @asubiotto for the backport
Reviewed 2 of 2 files at r1.
Reviewable status:complete! 1 of 0 LGTMs obtained
Contributor
Author
|
bors r=yuzefovich |
Contributor
Build failed (retrying...) |
Contributor
Author
|
bors r- |
Contributor
Canceled |
…led type Fixes cockroachdb#42994. Indexed columns are always decoded by the fetcher, even if they are unneeded. This PR adds the check to cfetcher initialization to error out if an indexed column is of an unneeded type. Release note (bug fix): Fixed a bug where scanning an index of an unsupported type with the vectorized engine would lead to an internal error.
5b8932b to
e6e9f0b
Compare
Contributor
Author
|
bors r=yuzefovich |
craig bot
pushed a commit
that referenced
this pull request
Dec 5, 2019
42250: storage/engine: introduce a datadriven framework to run MVCC tests r=itsbilal,petermattis a=knz
Previously MVCC tests were hand-coded in Go.
This was making it hard(er) to introduce new tests or modify existing
tests.
This commit improves upon this situation by introducing a new
datadriven test, `TestMVCCHistories`, which runs MVCC
tests written using a DSL:
```
begin_txn t=<name> [ts=<int>[,<int>]]
remove_txn t=<name>
resolve_intent t=<name> k=<key> [status=<txnstatus>]
restart_txn t=<name>
update_txn t=<name> t2=<name>
step_txn t=<name> [n=<int>]
advance_txn t=<name> ts=<int>[,<int>]
txn_status t=<name> status=<txnstatus>
check_intent k=<key> [none]
put [t=<name>] [ts=<int>[,<int>]] [resolve] k=<key> v=<string> [raw]
cput [t=<name>] [ts=<int>[,<int>]] [resolve] k=<key> v=<string> [raw] [cond=<string>]
increment [t=<name>] [ts=<int>[,<int>]] [resolve] k=<key> [inc=<val>]
del [t=<name>] [ts=<int>[,<int>]] [resolve] k=<key>
get [t=<name>] [ts=<int>[,<int>]] [resolve] k=<key> [inconsistent] [tombstones]
scan [t=<name>] [ts=<int>[,<int>]] [resolve] k=<key> [end=<key>] [inconsistent] [tombstones] [reverse]
merge [ts=<int>[,<int>]] k=<key> v=<string> [raw]
clear_range k=<key> end=<key>
```
Where `<key>` can be a simple string, or a string
prefixed by the following characters:
- `=foo` means exactly key `foo`
- `+foo` means `Key(foo).Next()`
- `-foo` means `Key(foo).PrefixEnd()`
Additionally, the pseudo-command `with` enables sharing
a group of arguments between multiple commands, for example:
```
with t=A
begin_txn
with k=a
put v=b
resolve_intent
```
Release note: None
42999: colexec: cfetcher errors on setup when indexed column contains unhandled type r=yuzefovich a=rohany
Fixes #42994.
Indexed columns are always decoded by the fetcher, even if they are unneeded.
This PR adds the check to cfetcher initialization to error out if an indexed
column is of an unneeded type.
Release note (bug fix): Fixed a bug where scanning an index of an unsupported
type with the vectorized engine would lead to an internal error.
Co-authored-by: Raphael 'kena' Poss <knz@thaumogen.net>
Co-authored-by: Rohan Yadav <rohany@alumni.cmu.edu>
Contributor
Build succeeded |
Contributor
Author
|
On second thought, I'm wondering if #43002 allows this test to pass without restricting the types of queries that the vectorized engine can run... |
Contributor
Author
|
I think we should revert this change if #43024 goes through. It is too restrictive and doesn't allow the cfetcher to skip columns in keys when decoding. |
Member
|
There was this failure with sqlsmith #42217 (comment). |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Fixes #42994.
Indexed columns are always decoded by the fetcher, even if they are unneeded.
This PR adds the check to cfetcher initialization to error out if an indexed
column is of an unneeded type.
Release note (bug fix): Fixed a bug where scanning an index of an unsupported
type with the vectorized engine would lead to an internal error.