[jdk17u-dev] RFR: 8267517: async logging for stdout and stderr

[Aleksey Shipilev]

On Fri, 7 Jul 2023 00:27:09 GMT, Xin Liu <xliu at openjdk.org> wrote:

> This PR backports JDK-8267517 to jdk17u. -Xlog:async will route LogOutput stdout/stderr to the asynchronous logging queue like ordinary log files. The reason we do is that some Java application outputs logs to stdout/stderr and redirect them to files then. 
> 
> This is **NOT** a clean backport. It's because jdk17u is missing a few patches, such as JDK-8273471, JDK-8269004. I tried to backport them beforehand, but they both have dependencies.  I manually resolve a couple of conflicts. 
> 
> We have verified patch like how we verified for jdk-tip. We output verbose gclog to stdout `-Xlog:gc=trace:stdout` and leverage software flow control to pause our console(C-s). Java threads don't be blocked with '-Xlog:async'. Without this patch, kernel puts Java threads which require GC to 'sleep' state when unified log is blocked.
> 
> ---------
> ### Progress
> - [ ] Change must be properly reviewed (1 review required, with at least 1 [Reviewer](https://openjdk.org/bylaws#reviewer))
> - [x] Change must not contain extraneous whitespace
> - [x] Commit message must refer to an issue
> 
> ### Error
>  ⚠️ The pull request body must not be empty.
> 
> 
> 
> ### Reviewing
> <details><summary>Using <code>git</code></summary>
> 
> Checkout this PR locally: \
> `$ git fetch https://git.openjdk.org/jdk17u-dev.git pull/1555/head:pull/1555` \
> `$ git checkout pull/1555`
> 
> Update a local copy of the PR: \
> `$ git checkout pull/1555` \
> `$ git pull https://git.openjdk.org/jdk17u-dev.git pull/1555/head`
> 
> </details>
> <details><summary>Using Skara CLI tools</summary>
> 
> Checkout this PR locally: \
> `$ git pr checkout 1555`
> 
> View PR using the GUI difftool: \
> `$ git pr show -t 1555`
> 
> </details>
> <details><summary>Using diff file</summary>
> 
> Download this PR as a diff file: \
> <a href="https://git.openjdk.org/jdk17u-dev/pull/1555.diff">https://git.openjdk.org/jdk17u-dev/pull/1555.diff</a>
> 
> </details>

Looks okay with minor nit.

Please also run `tier1 tier2 tier3` on any convenient architecture for this.

src/hotspot/share/logging/logFileStreamOutput.cpp line 124:

> 122:   const bool use_decorations = !_decorators.is_empty();
> 123: 
> 124:   int written = 0;

Since this is unclean backport anyway, is there any value to match the order of these statements to the order upstream? Look how it is in `openjdk/jdk`: https://github.com/openjdk/jdk/blob/06a1a15d014f5ca48f62f5f0c8e8682086c4ae0b/src/hotspot/share/logging/logFileStreamOutput.cpp#L118-L120

-------------

Marked as reviewed by shade (Reviewer).

PR Review: https://git.openjdk.org/jdk17u-dev/pull/1555#pullrequestreview-1521639467
PR Comment: https://git.openjdk.org/jdk17u-dev/pull/1555#issuecomment-1628595534
PR Review Comment: https://git.openjdk.org/jdk17u-dev/pull/1555#discussion_r1257966155