Re: RFR: 8295857: Clarify that cleanup code can be skipped when the JVM terminates (e.g. when calling halt()) [v4]
On Tue, 22 Nov 2022 00:50:51 GMT, Brent Christian wrote:
>> [JDK-8290036](https://bugs.openjdk.org/browse/JDK-8290036) documented the
>> shutdown sequence, noting that calling Runtime.halt() skips the shutdown
>> sequence and immediately terminates the VM. Thus, "threads' current methods
>> do not complete normally or abruptly; no finally clause of any method is
>> executed".
>>
>> One ramification of this is that resources within try-with-resource blocks
>> will not be released. It would be good to state this explicitly.
>
> Brent Christian has updated the pull request incrementally with one
> additional commit since the last revision:
>
> Update Runtime class doc re: other unexpected behaviors
src/java.base/share/classes/java/lang/Runtime.java line 285:
> 283: * actions intended to be performed by shutdown hooks,
> 284: * {@linkplain Thread.UncaughtExceptionHandler uncaught exception
> handlers},
> 285: * finally blocks, or try-with-resources blocks. This can lead to
> data corruption.
Since the examples of the consequences of halting are covered above in "JVM
Termination" and are getting a bit involved, I'd suggest not trying to list the
concrete consequences here, and referring the reader (again) to the "JVM
Termination" section.
-
PR: https://git.openjdk.org/jdk/pull/11218
Re: RFR: 8295857: Clarify that cleanup code can be skipped when the JVM terminates (e.g. when calling halt()) [v4]
On Tue, 22 Nov 2022 22:24:00 GMT, David Holmes wrote:
>> Brent Christian has updated the pull request incrementally with one
>> additional commit since the last revision:
>>
>> Update Runtime class doc re: other unexpected behaviors
>
> src/java.base/share/classes/java/lang/Runtime.java line 97:
>
>> 95: * no {@code finally} clause of any method is executed, and
>> try-with-resources
>> 96: * blocks do not {@linkplain AutoCloseable close} their resources.
>> 97: * Other unexpected behaviors may also result.
>
> No sorry that doesn't work for me. It sounds like you might get unexpected
> things happening, when in fact it is things not happening that might surprise
> someone (who doesn't understand what "immediately prevented from executing
> any further Java code" means). The list of examples is getting too long to
> handle in sentence structure so I suggest an actual list (not sure of list
> syntax). Suggestion:
>
>
> * Java code. This includes shutdown hooks as well as daemon and non-daemon
> threads. This
> * means, for example, that:
> * - threads' current methods do not complete normally or abruptly
> * - no {@code finally} clause of any method is executed
> * - no {@linkplain Thread.UncaughtExceptionHandler uncaught exception
> handler} is executed
> * - no try-with-resources blocks {@linkplain AutoCloseable close} their
> resources
I think it's reasonable to position these as examples and to put them into list
form.
(wordsmithing) For parallel structure I'd suggest "X does not occur" instead of
"no X occurs" in all but the first item, specifically (markup elided):
- threads' current methods do not complete normally or abruptly;
- finally clauses are not executed;
- uncaught exception handlers are not executed; and
- resources opened by try-with-resources statements are not closed.
-
PR: https://git.openjdk.org/jdk/pull/11218
Re: RFR: 8295857: Clarify that cleanup code can be skipped when the JVM terminates (e.g. when calling halt()) [v4]
On Tue, 22 Nov 2022 00:50:51 GMT, Brent Christian wrote:
>> [JDK-8290036](https://bugs.openjdk.org/browse/JDK-8290036) documented the
>> shutdown sequence, noting that calling Runtime.halt() skips the shutdown
>> sequence and immediately terminates the VM. Thus, "threads' current methods
>> do not complete normally or abruptly; no finally clause of any method is
>> executed".
>>
>> One ramification of this is that resources within try-with-resource blocks
>> will not be released. It would be good to state this explicitly.
>
> Brent Christian has updated the pull request incrementally with one
> additional commit since the last revision:
>
> Update Runtime class doc re: other unexpected behaviors
src/java.base/share/classes/java/lang/Runtime.java line 97:
> 95: * no {@code finally} clause of any method is executed, and
> try-with-resources
> 96: * blocks do not {@linkplain AutoCloseable close} their resources.
> 97: * Other unexpected behaviors may also result.
No sorry that doesn't work for me. It sounds like you might get unexpected
things happening, when in fact it is things not happening that might surprise
someone (who doesn't understand what "immediately prevented from executing any
further Java code" means). The list of examples is getting too long to handle
in sentence structure so I suggest an actual list (not sure of list syntax).
Suggestion:
* Java code. This includes shutdown hooks as well as daemon and non-daemon
threads. This
* means, for example, that:
* - threads' current methods do not complete normally or abruptly
* - no {@code finally} clause of any method is executed
* - no {@linkplain Thread.UncaughtExceptionHandler uncaught exception
handler} is executed
* - no try-with-resources blocks {@linkplain AutoCloseable close} their
resources
-
PR: https://git.openjdk.org/jdk/pull/11218
Re: RFR: 8295857: Clarify that cleanup code can be skipped when the JVM terminates (e.g. when calling halt()) [v4]
On Tue, 22 Nov 2022 00:50:51 GMT, Brent Christian wrote: >> [JDK-8290036](https://bugs.openjdk.org/browse/JDK-8290036) documented the >> shutdown sequence, noting that calling Runtime.halt() skips the shutdown >> sequence and immediately terminates the VM. Thus, "threads' current methods >> do not complete normally or abruptly; no finally clause of any method is >> executed". >> >> One ramification of this is that resources within try-with-resource blocks >> will not be released. It would be good to state this explicitly. > > Brent Christian has updated the pull request incrementally with one > additional commit since the last revision: > > Update Runtime class doc re: other unexpected behaviors Marked as reviewed by lancea (Reviewer). - PR: https://git.openjdk.org/jdk/pull/11218
Re: RFR: 8295857: Clarify that cleanup code can be skipped when the JVM terminates (e.g. when calling halt()) [v4]
> [JDK-8290036](https://bugs.openjdk.org/browse/JDK-8290036) documented the > shutdown sequence, noting that calling Runtime.halt() skips the shutdown > sequence and immediately terminates the VM. Thus, "threads' current methods > do not complete normally or abruptly; no finally clause of any method is > executed". > > One ramification of this is that resources within try-with-resource blocks > will not be released. It would be good to state this explicitly. Brent Christian has updated the pull request incrementally with one additional commit since the last revision: Update Runtime class doc re: other unexpected behaviors - Changes: - all: https://git.openjdk.org/jdk/pull/11218/files - new: https://git.openjdk.org/jdk/pull/11218/files/cbba781f..6df8acf8 Webrevs: - full: https://webrevs.openjdk.org/?repo=jdk=11218=03 - incr: https://webrevs.openjdk.org/?repo=jdk=11218=02-03 Stats: 1 line in 1 file changed: 1 ins; 0 del; 0 mod Patch: https://git.openjdk.org/jdk/pull/11218.diff Fetch: git fetch https://git.openjdk.org/jdk pull/11218/head:pull/11218 PR: https://git.openjdk.org/jdk/pull/11218
