[SPARK-17618] Fix invalid comparisons between UnsafeRow and other row formats #15185
Conversation
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
|
Test build #65733 has finished for PR 15185 at commit
|
|
LGTM |
|
Test build #65734 has finished for PR 15185 at commit
|
|
Jenkins, retest this please. |
|
Test build #65741 has finished for PR 15185 at commit
|
|
Jenkins, retest this please. |
|
Test build #65750 has finished for PR 15185 at commit
|
asfgit
pushed a commit
that referenced
this pull request
Sep 27, 2016
… formats ## What changes were proposed in this pull request? This patch addresses a correctness bug in Spark 1.6.x in where `coalesce()` declares that it can process `UnsafeRows` but mis-declares that it always outputs safe rows. If UnsafeRow and other Row types are compared for equality then we will get spurious `false` comparisons, leading to wrong answers in operators which perform whole-row comparison (such as `distinct()` or `except()`). An example of a query impacted by this bug is given in the [JIRA ticket](https://issues.apache.org/jira/browse/SPARK-17618). The problem is that the validity of our row format conversion rules depends on operators which handle `unsafeRows` (signalled by overriding `canProcessUnsafeRows`) correctly reporting their output row format (which is done by overriding `outputsUnsafeRows`). In #9024, we overrode `canProcessUnsafeRows` but forgot to override `outputsUnsafeRows`, leading to the incorrect `equals()` comparison. Our interface design is flawed because correctness depends on operators correctly overriding multiple methods this problem could have been prevented by a design which coupled row format methods / metadata into a single method / class so that all three methods had to be overridden at the same time. This patch addresses this issue by adding missing `outputsUnsafeRows` overrides. In order to ensure that bugs in this logic are uncovered sooner, I have modified `UnsafeRow.equals()` to throw an `IllegalArgumentException` if it is called with an object that is not an `UnsafeRow`. ## How was this patch tested? I believe that the stronger misuse-checking in `UnsafeRow.equals()` is sufficient to detect and prevent this class of bug. Author: Josh Rosen <joshrosen@databricks.com> Closes #15185 from JoshRosen/SPARK-17618.
|
Merging to master. Thanks! |
asfgit
pushed a commit
that referenced
this pull request
Sep 27, 2016
… other formats This patch ports changes from #15185 to Spark 2.x. In that patch, a correctness bug in Spark 1.6.x which was caused by an invalid `equals()` comparison between an `UnsafeRow` and another row of a different format. Spark 2.x is not affected by that specific correctness bug but it can still reap the error-prevention benefits of that patch's changes, which modify ``UnsafeRow.equals()` to throw an IllegalArgumentException if it is called with an object that is not an `UnsafeRow`. Author: Josh Rosen <joshrosen@databricks.com> Closes #15265 from JoshRosen/SPARK-17618-master.
asfgit
pushed a commit
that referenced
this pull request
Sep 27, 2016
… other formats This patch ports changes from #15185 to Spark 2.x. In that patch, a correctness bug in Spark 1.6.x which was caused by an invalid `equals()` comparison between an `UnsafeRow` and another row of a different format. Spark 2.x is not affected by that specific correctness bug but it can still reap the error-prevention benefits of that patch's changes, which modify ``UnsafeRow.equals()` to throw an IllegalArgumentException if it is called with an object that is not an `UnsafeRow`. Author: Josh Rosen <joshrosen@databricks.com> Closes #15265 from JoshRosen/SPARK-17618-master. (cherry picked from commit 2f84a68) Signed-off-by: Josh Rosen <joshrosen@databricks.com>
zzcclp
pushed a commit
to zzcclp/spark
that referenced
this pull request
Sep 28, 2016
… formats ## What changes were proposed in this pull request? This patch addresses a correctness bug in Spark 1.6.x in where `coalesce()` declares that it can process `UnsafeRows` but mis-declares that it always outputs safe rows. If UnsafeRow and other Row types are compared for equality then we will get spurious `false` comparisons, leading to wrong answers in operators which perform whole-row comparison (such as `distinct()` or `except()`). An example of a query impacted by this bug is given in the [JIRA ticket](https://issues.apache.org/jira/browse/SPARK-17618). The problem is that the validity of our row format conversion rules depends on operators which handle `unsafeRows` (signalled by overriding `canProcessUnsafeRows`) correctly reporting their output row format (which is done by overriding `outputsUnsafeRows`). In apache#9024, we overrode `canProcessUnsafeRows` but forgot to override `outputsUnsafeRows`, leading to the incorrect `equals()` comparison. Our interface design is flawed because correctness depends on operators correctly overriding multiple methods this problem could have been prevented by a design which coupled row format methods / metadata into a single method / class so that all three methods had to be overridden at the same time. This patch addresses this issue by adding missing `outputsUnsafeRows` overrides. In order to ensure that bugs in this logic are uncovered sooner, I have modified `UnsafeRow.equals()` to throw an `IllegalArgumentException` if it is called with an object that is not an `UnsafeRow`. ## How was this patch tested? I believe that the stronger misuse-checking in `UnsafeRow.equals()` is sufficient to detect and prevent this class of bug. Author: Josh Rosen <joshrosen@databricks.com> Closes apache#15185 from JoshRosen/SPARK-17618. (cherry picked from commit e2ce0ca)
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
What changes were proposed in this pull request?
This patch addresses a correctness bug in Spark 1.6.x in where
coalesce()declares that it can processUnsafeRowsbut mis-declares that it always outputs safe rows. If UnsafeRow and other Row types are compared for equality then we will get spuriousfalsecomparisons, leading to wrong answers in operators which perform whole-row comparison (such asdistinct()orexcept()). An example of a query impacted by this bug is given in the JIRA ticket.The problem is that the validity of our row format conversion rules depends on operators which handle
unsafeRows(signalled by overridingcanProcessUnsafeRows) correctly reporting their output row format (which is done by overridingoutputsUnsafeRows). In #9024, we overrodecanProcessUnsafeRowsbut forgot to overrideoutputsUnsafeRows, leading to the incorrectequals()comparison.Our interface design is flawed because correctness depends on operators correctly overriding multiple methods this problem could have been prevented by a design which coupled row format methods / metadata into a single method / class so that all three methods had to be overridden at the same time.
This patch addresses this issue by adding missing
outputsUnsafeRowsoverrides. In order to ensure that bugs in this logic are uncovered sooner, I have modifiedUnsafeRow.equals()to throw anIllegalArgumentExceptionif it is called with an object that is not anUnsafeRow.How was this patch tested?
I believe that the stronger misuse-checking in
UnsafeRow.equals()is sufficient to detect and prevent this class of bug.